WO2012060290A1 - Image processing apparatus, image processing method, and display apparatus - Google Patents

Image processing apparatus, image processing method, and display apparatus Download PDF

Info

Publication number
WO2012060290A1
WO2012060290A1 PCT/JP2011/074921 JP2011074921W WO2012060290A1 WO 2012060290 A1 WO2012060290 A1 WO 2012060290A1 JP 2011074921 W JP2011074921 W JP 2011074921W WO 2012060290 A1 WO2012060290 A1 WO 2012060290A1
Authority
WO
WIPO (PCT)
Prior art keywords
pixel
noise
value
processing target
pixels
Prior art date
Application number
PCT/JP2011/074921
Other languages
French (fr)
Japanese (ja)
Inventor
沼尾 孝次
崇志 峰
合志 清一
Original Assignee
シャープ株式会社
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by シャープ株式会社 filed Critical シャープ株式会社
Publication of WO2012060290A1 publication Critical patent/WO2012060290A1/en

Links

Images

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N5/00Details of television systems
    • H04N5/14Picture signal circuitry for video frequency region
    • H04N5/21Circuitry for suppressing or minimising disturbance, e.g. moiré or halo
    • H04N5/213Circuitry for suppressing or minimising impulsive noise

Definitions

  • the present invention relates to an image processing device, an image processing method, and a display device.
  • noise components such as so-called snow noise, Gaussian noise, and shot noise are mixed in an image signal when an image is captured and transmitted.
  • noise components such as so-called snow noise, Gaussian noise, and shot noise are mixed in an image signal when an image is captured and transmitted.
  • the received signal that has received the broadcast wave becomes a video signal with a lot of noise components.
  • analog video recorded in the past is often digitized and rebroadcast, and the video signal in that case is a video signal with a lot of noise components.
  • a low-pass filter As a technique for reducing this noise component, a low-pass filter is known.
  • the pixel value for example, the value of a luminance signal
  • the noise reduction circuit performs noise reduction processing to reduce the noise, thereby improving the image quality.
  • the image quality deteriorates at that point due to blurring of the edge.
  • an adaptive low-pass filter using the median filter shown below has been proposed.
  • a noise reduction circuit detects a noise component of a line portion having no image signal during a vertical blanking period of the video signal, and noise in the video signal is based on the voltage (noise amount E3) of the noise component. It has been shown to reduce. Specifically, as shown in FIG. 19, the noise reduction circuit compares the processing pixel voltage E1 from the delay circuit 512 with the output voltage E2 after the median filter processing is performed in the median filter 513.
  • FIG. 20 is a diagram for explaining the result of the noise reduction processing of the noise reduction circuit in Patent Document 1.
  • FIG. 20 is a diagram for explaining the result of the noise reduction processing of the noise reduction circuit in Patent Document 1.
  • Patent Document 2 describes that noise removal processing is performed using a noise removal device 601 having the configuration shown in FIG. 21 as an adaptive low-pass filter. That is, the subtracting unit 612 obtains a difference Diff one frame before from the signal (1 frame delayed signal) Im1 after noise removal and the input signal Di0. The motion / noise detection unit 614 determines whether the difference Diff is due to image motion or noise, and outputs the determination result as a motion degree signal MDS. The coefficient conversion means 616 converts the motion degree signal MDS into a cyclic coefficient Km.
  • the multiplying unit 617 multiplies the output Dfn whose amplitude limiting unit 613 limits the amplitude of the difference Diff by a cyclic coefficient Km to obtain a noise cyclic amount Nd.
  • the adding means 618 outputs the output signal Do0 as an image signal after filtering by adding the noise circulation amount Nd to the input signal Di0.
  • the noise removal apparatus 601 uses the output signal Do0 as a signal Im1 after noise removal for one-frame delayed filtering of the current image signal to be processed. For this reason, the pixel difference between frames is discriminated as image motion or noise, and noise is removed by setting a coefficient for noise according to the motion degree signal MDS. Therefore, processing for removing image motion as noise is performed. Can be reduced.
  • FIG. 22 is a diagram illustrating an original image obtained by artificially adding discrete noise to captured image data.
  • FIG. 23 is a diagram showing a result of performing noise removal processing on the image of FIG. 22 using the noise reduction circuit of Patent Document 1. The image shown in FIG. 23 is obtained by performing noise removal processing once in the horizontal direction (direction parallel to the X axis) on the original image shown in FIG. Many remain on the face, indicating that noise removal is insufficient.
  • FIG. 24 is a diagram illustrating a result of performing noise processing in the vertical direction (direction parallel to the Y axis) on the image of FIG. 23 using the noise reduction circuit of Patent Document 1.
  • the image of FIG. 24 has the noise removal process sufficiently performed by removing the discrete noise of the female face portion as compared with the image of FIG.
  • the discrete noise is sufficiently removed.
  • the image of FIG. It can be seen that the hair on the face and the outline of the leaf of the plant on the left are quite blurred.
  • FIG. 25 shows No. in the standard moving image of ITE (The Institute of Image Information and Television Engineers). A part of the image of 32 “moss and stone Buddha” is shown.
  • FIG. 26 shows an image obtained by performing noise reduction processing in Patent Document 1 on the image of FIG.
  • the dent that can be recognized due to the difference in color within the region denoted by reference numeral R1 is blurred and disappears in the image of FIG. I know that.
  • the noise in the image of FIG. 22 or the image of FIG. 25 is deleted, the degree of blur increases and the outline disappears as seen in the image of FIG. 24 or the image of FIG. It becomes.
  • the present invention has been made in view of the above problems, and an image processing apparatus, an image processing method, and an image processing apparatus that can eliminate noise with reduced blurring and reduced blur are used. It is to provide a display device.
  • an image processing apparatus is an object of noise removal processing in a pixel space formed from pixels of a plurality of frames included in a video signal.
  • a noise determination unit that determines whether the pixel value of the processing target pixel belongs to an upper group, a middle group, or a lower group; and when the processing target pixel belongs to the upper group, the pixel value of the processing target pixel
  • the noise amount is calculated from the pixel value of the processing target pixel. Summed, if the processing target pixel belongs to the middle group, and having a noise reduction operation unit, not subtracting the amount of noise to the pixel value of the processing target pixel.
  • the number of pixels included in each of the upper group, the middle group, and the lower group is set according to a type of video signal, and the noise determination unit is configured to compare the comparison pixel group.
  • the pixel values of at least three pixels are ranked in descending order and are distributed according to the number of pixels set in each of the upper group, the middle group, and the lower group.
  • a line connecting the processing target pixel and a pixel around the processing target pixel is a part of a contour line included in the frame.
  • a pixel around the processing target pixel in the configuration is selected as the comparison pixel.
  • the image processing apparatus is characterized in that the comparison pixel is selected from the pixels set in advance in a horizontal direction, a vertical direction, and an oblique direction with respect to the processing target pixel.
  • the comparison pixel is rotated at a preset angle clockwise in the x-axis direction, the y-axis direction, and the origin in the clockwise direction with the processing target pixel in the pixel space as the origin.
  • the pixel is selected from the pixels in the axial direction, which is rotated in the axial direction by a preset angle counterclockwise around the origin.
  • An image processing apparatus is characterized in that the pixel space includes a pixel array in a three-dimensional space with a time direction added.
  • An image processing apparatus performs first-order noise processing when the pixel space includes a pixel array in a three-dimensional space with a time direction added, and the pixel space includes the processing target pixel. In the case of a pixel array in a two-dimensional space, second-order noise processing is performed.
  • the image processing apparatus further includes a noise value detection circuit that calculates the preset amount of noise, and the noise value detection circuit includes the processing target pixel and the processing target in the pixel space.
  • the noise amount is calculated based on a difference from a plurality of other pixels around the pixel.
  • the noise reduction calculation unit obtains a middle group average value that is an average value of pixel values of the middle group pixels, and the processing target pixel is the upper group or the When belonging to a lower group, the middle group average value is subtracted from the pixel value of the processing target pixel, and the noise value detection circuit obtains an absolute value as a result of subtraction by the noise reduction calculation unit, and the two-dimensional array
  • the noise amount is obtained by averaging the absolute value obtained for each of the processing target pixels selected in step 2 in the two-dimensional array.
  • the noise reduction calculation unit determines that the pixel value of the processing target pixel is the pixel value of the processing target pixel and It is determined whether or not the pixel value of the plurality of comparison pixels is the maximum value or the minimum value. If the value is the maximum value or the minimum value, the difference between the pixel value of the processing target pixel and each comparison pixel A difference average value which is an average value is calculated and output to the noise detection unit, and the noise value detection circuit calculates an absolute value of the difference average value calculated by the noise reduction calculation unit, and the calculated difference average The noise amount is obtained by averaging absolute values of the values in the pixel space.
  • the noise reduction calculation unit subtracts the middle group average value from the pixel value of the processing target pixel.
  • the noise value detection circuit calculates an absolute value of the difference average value calculated by the noise reduction calculation unit, and averages the absolute value of the difference average value in the pixel space. The amount of noise is obtained by the following.
  • the noise reduction calculation unit may determine whether the pixel value of the processing target pixel is the maximum value or the minimum value when the processing target pixel belongs to the upper group or the lower group. If it is a maximum value or a minimum value, a difference average value that is an average value of differences between the pixel value of the processing target pixel and the comparison pixel is calculated, and the noise value detection circuit The absolute value of the difference average value calculated by the noise reduction calculation unit is obtained, a histogram is generated for each numerical value of the difference average value for each frame, and the mode value of the generated histogram is used as the noise amount. And
  • the noise determination unit includes a processing target pixel that is a target of noise removal processing in a pixel space formed from pixels of a plurality of frames included in a video signal, and the pixel space includes the processing target pixel.
  • a comparison pixel group with at least three comparison pixels located around the processing target pixel, comparing pixel values of a plurality of pixels belonging to the comparison pixel group, and the pixel value of the processing target pixel is an upper group
  • the noise determination process for determining whether it belongs to the middle group or the lower group, and the noise reduction calculation unit are preset from the pixel value of the processing target pixel when the processing target pixel belongs to the upper group.
  • the noise amount is added from the pixel value of the processing target pixel. If the processing target pixel belongs to the middle group, characterized in that it comprises a noise reduction operation process without subtracting the amount of noise to the pixel value of the target pixel, a.
  • An image processing program provides a computer with a processing target pixel to be subjected to noise removal processing in a pixel space formed from pixels of a plurality of frames included in a video signal, and the processing target in the pixel space.
  • a comparison pixel group is configured from at least three comparison pixels located around the pixel, and pixel values of a plurality of pixels belonging to the comparison pixel group are compared.
  • the display device of the present invention includes any one of the above image processing devices, and performs video display after removing noise from the video signal.
  • the present invention in the noise removal of an image, it is possible to reduce noise by reducing the elimination of the outline of the image and suppressing blurring.
  • FIG. 3 is a functional block diagram of a liquid crystal display device according to first to third embodiments of the present invention. It is a figure which shows the connection relation of the signal of the liquid crystal drive part 15 and the liquid crystal panel 16 in FIG. It is a figure for demonstrating the outline
  • FIG. 1 It is a figure of the result of having performed the noise removal by 2nd Embodiment with respect to the original image of FIG. It is a figure which shows the structural example of the noise reduction part 13 by 3rd Embodiment. It is a figure which shows the structural example of 13 A of 1st noise reduction parts in 3rd Embodiment. It is a figure which shows the structural example of the 2nd noise reduction part 13B in 3rd Embodiment. It is a figure of the result of having performed the noise removal by this embodiment with respect to the original image of FIG. It is a figure which shows the structure of the noise reduction circuit in patent document 1. FIG. It is a figure explaining the result of the noise reduction process of the noise reduction circuit in patent document 1. FIG.
  • FIG. 22 shows the structure of the adaptive low-pass filter of the noise removal apparatus in patent document 2. It is a figure which shows the original image which added discrete noise artificially with respect to the imaged image data. It is a figure which shows the image of the result of having performed the noise removal process with respect to the image of FIG. 22 using the noise reduction circuit of patent document 1.
  • FIG. 23 In contrast to FIG. 23 in which noise reduction processing in the horizontal direction is performed on the image in FIG. 22 using the noise reduction circuit in Patent Document 1, the noise reduction circuit in Patent Document 1 is used in the vertical direction with respect to FIG. It is a figure which shows the result of having performed the noise process. No. in the ITE standard moving image. A part of the image of 32 “moss and stone Buddha” is shown. The image obtained by performing the noise reduction process in patent document 1 is shown with respect to the image of FIG.
  • FIG. 1 is a functional block diagram of a liquid crystal display device according to the first embodiment of the present invention.
  • the liquid crystal display device 1 includes a detection unit 11, a Y / C (luminance signal / color signal) separation unit 12, a noise reduction unit 13, and an RGB (Red: red, Green: green, Blue: blue) conversion unit 14.
  • the liquid crystal drive unit 15 and the liquid crystal panel 16 are provided.
  • the detection unit 11 receives high-frequency signals of image data of a plurality of channels of terrestrial analog television broadcast supplied from an antenna (not shown). Then, the detection unit 11 extracts a modulation signal of a desired channel from the high-frequency signal supplied from the antenna, converts the extracted modulation signal into a baseband signal, and outputs it to the Y / C separation unit 12.
  • the Y / C separation unit 12 demodulates the supplied baseband signal, separates it into a luminance signal Y, a color difference signal Cb, and a color difference signal Cr, and converts each separated signal into a digital signal at a predetermined sampling frequency. Convert to The Y / C separation unit 12 also arranges the luminance signal Y converted into a digital signal, the color difference signal Cb, and the color difference signal Cr for each pixel arranged adjacent to each other in the main scanning direction (horizontal direction, horizontal direction) of the image. Synchronize and output to the noise reduction unit 13. Each of the luminance signal Y, the color difference signal Cb, and the color difference signal Cr is a numerical value indicated by a digital signal indicating a signal intensity expressed by a voltage value.
  • the noise reduction unit 13 compares the supplied luminance signal Y, color difference signal Cb, and color difference signal Cr between pixels in the same frame (pixel space in which pixels are arranged), and noise is superimposed on the processing target pixel. It is determined whether or not.
  • the noise reduction unit 13 calculates the amount of noise in frame units or field units.
  • the noise reduction unit 13 performs noise removal processing on the processing target pixel by adding and subtracting the noise amount calculated from the processing target pixel determined to have noise superimposed thereon, and the processed luminance signal Y and color difference signal Cb and the color difference signal Cr are synchronized and output to the RGB converter 14 as a video signal. Details of the noise processing for each pixel in the noise reduction unit 13 will be described later.
  • the video signal is interlaced
  • noise processing is performed for each field.
  • noise processing is performed for each frame.
  • the RGB conversion unit 14 converts it into a progressive signal. In addition, the RGB conversion unit 14 adjusts the number of pixels (scaling processing) in accordance with the resolution of the liquid crystal panel 16 with respect to the progressive signal. Then, the RGB converter 14 converts the video signal with the adjusted number of pixels into an RGB signal (Red, Green, Blue color video signal), and outputs the converted RGB signal to the liquid crystal driver 15.
  • RGB signal Red, Green, Blue color video signal
  • the liquid crystal drive unit 15 generates a clock signal or the like for displaying video data supplied to the liquid crystal panel 16 on a two-dimensional plane of the screen. Then, the liquid crystal drive unit 15 supplies the generated clock signal to the liquid crystal panel 16.
  • FIG. 2 is a diagram illustrating a signal connection relationship between the liquid crystal driving unit 15 and the liquid crystal panel 16.
  • the liquid crystal driving unit 15 includes a source driver unit and a gate driver unit.
  • a display arranged at a point where the source line 19 and the gate line 18 intersect with each other.
  • the element PIX that is, the display element PIX arranged on the matrix is controlled to display an image.
  • the display element PIX includes a TFT (Thin Film Transistor) and a liquid crystal pixel element into which a voltage corresponding to a gradation described later is written by the TFT.
  • TFT Thin Film Transistor
  • the source driver unit holds the supplied RGB signals (for example, digital signals indicating the respective RGB pixel values) for each source line 19 (wiring in the column direction) of the liquid crystal panel 16 by a hold circuit provided therein. To do.
  • the source driver unit generates a voltage (source signal) that is grayscaled for pixel element driving from the held RGB signal in synchronization with the gate clock signal for the vertical arrangement of the screen.
  • the source signal thus supplied is supplied to the source line 19 of the TFT connected to the liquid crystal element PIX of the liquid crystal panel 16.
  • the gate driver unit supplies a gate clock signal to one row of the liquid crystal elements PIX on the screen through TFT gate lines 18 (corresponding to horizontal wiring, main scanning) connected to the liquid crystal elements PIX of the liquid crystal panel 16.
  • a predetermined gate signal is supplied in synchronization with the above.
  • the liquid crystal panel 16 includes an array substrate, a counter substrate, and a liquid crystal. At each intersection of the source line 19 and the gate line 18 on the array substrate, a liquid crystal element PIX, that is, a pixel electrode connected to the TFT and the drain electrode of the TFT, and a counter electrode (consisting of a solid electrode on the counter substrate) ) Are arranged one by one. Here, in the pixel element, liquid crystal is sealed between the pixel electrode and the counter electrode.
  • the liquid crystal panel 16 corresponds to the three primary colors RGB (Red, Green, Blue) for each pixel, that is, for every three liquid crystal elements PIX.
  • the liquid crystal panel 16 has one TFT for each subpixel.
  • the TFT is selected and turned on when the gate signal supplied from the gate driver unit is supplied to the gate electrode and the gate signal is at a high level, for example.
  • the source signal supplied from the source driver is supplied to the source electrode of the TFT, and when the TFT is in an ON state, a gradation voltage is applied to the pixel electrode connected to the drain electrode of the TFT, that is, the pixel element. Is done.
  • the orientation of the liquid crystal of the pixel element changes in accordance with the gradation voltage, thereby changing the light transmittance of the liquid crystal in the region of the pixel element.
  • the gradation voltage is held in the liquid crystal capacitance of the pixel element formed by the liquid crystal portion between the pixel electrode connected to the drain electrode of the TFT and the counter electrode, and the alignment of the liquid crystal is maintained.
  • the alignment of the liquid crystal is maintained until the next signal is supplied to the source electrode and the maintained voltage value is changed by the gradation voltage, so that the light transmittance of the liquid crystal is maintained.
  • the liquid crystal panel 16 performs gradation display on the supplied video data.
  • the transmissive liquid crystal panel has been described here, the present invention is not limited to this, and a reflective liquid crystal panel may be used.
  • FIG. 3 is a diagram for explaining an outline of noise removal in the present embodiment.
  • V indicates the vertical direction (Y-axis method)
  • H indicates the horizontal direction (X-axis direction)
  • F indicates the frame direction (time-axis direction).
  • (Fi, xj, yk) is Fi in time sequence of frames
  • xj is a source line number
  • yk is a gate line (main scanning line) number
  • Fi, xj and yk are integers.
  • the arrangement of pixels in the pixel space (the same applies to FIG. 10 described later).
  • the frame in which the processing target pixel S0 exists is the F0 frame, and in this F0 frame, the x2th source line and the y2th gate line The position of the pixel of the pixel element PIX at the intersection is shown.
  • the noise reduction unit 13 when the noise reduction unit 13 performs noise removal, a process for determining whether noise is superimposed on a pixel to be processed, that is, a pixel to be processed will be described.
  • the noise reduction unit 13 is supplied with a luminance signal and a color signal from the Y / C separation unit 12 in the horizontal direction, that is, in the main scanning direction (direction of the gate line 18) in units of pixels in order to the liquid crystal elements of the liquid crystal panel 16. .
  • the pixels of the pixels arranged on the second row main scanning line are completed.
  • Video signals are input, and video signals corresponding to sequential main scanning lines are input. Further, when the video signal for one frame is supplied to the noise reduction unit 13, the video signal of the next frame is sequentially supplied.
  • the noise reduction unit 13 compares pixel values between pixels arranged in a two-dimensional plane (two-dimensional space) of one frame, and determines whether noise is superimposed on the pixels. .
  • the noise reduction unit 13 performs noise removal processing independently on each of the luminance signal Y and the color signal (color difference signal Cb, color difference signal Cr). Will be described.
  • the noise reduction unit 13 forms a comparison pixel group by nine pixels including a processing target pixel S0 and pixels S1 to S8 (comparison pixels) around the processing target pixel S0.
  • the noise reduction unit 13 determines the size of the pixel value of the processing target pixel S0 and each of the comparison pixels S1 to S8 in the nine pixels in the comparison pixel group, and determines the processing target from the determination result of the size relationship. Each of the number of pixels having a pixel value larger than the pixel S0 and the number of pixels smaller than the processing target pixel S0 are counted.
  • the noise reduction unit 13 uses the number of comparison pixels having a pixel value larger than that of the processing target pixel S0 and the number of comparison pixels having a pixel value smaller than that of the processing target pixel S0,
  • the group is divided into three groups, a rank group and a subordinate group. That is, when the number of comparison pixels having a pixel value larger than that of the processing target pixel S0 is smaller than the preset number of pixels as the upper group, the noise reduction unit 13 sets the processing target pixel S0 as the upper group.
  • the noise reduction unit 13 sets the processing target pixel S0 as the lower group when the number of comparison pixels having a pixel value smaller than that of the processing target pixel S0 is smaller than the preset number of pixels as the lower group.
  • the noise reduction unit 13 has a larger number of comparison pixels having a pixel value larger than the processing target pixel S0 than the preset number of pixels as the upper group, and a processing target from the preset number of pixels as the lower group.
  • the processing target pixel S0 is set as a middle group.
  • the noise reduction unit 13 determines that noise is superimposed on the processing target pixel S0.
  • the processing target pixel S0 belongs to the middle group, the noise reduction unit 13 It is determined that no noise is superimposed on the.
  • the noise reduction unit 13 sets the upper two pixels (first to second pixels from the top to the second) in the above-described order as an upper group, and the lower two in the order array.
  • the pixels (9th to 8th pixels from the lowest to the second) are set as the lower group, and the remaining 5 pixels (3rd to 7th pixels) are set as the middle group.
  • FIG. 4 is a diagram showing a correspondence relationship between each pixel in a two-dimensional plane frame and a contour line of a figure. The reason why the number of pixels in the upper group and the number of pixels in the lower group, two upper groups, and two lower groups in this embodiment will be described with reference to FIG. As shown in FIG.
  • the contour lines of the figure in the image are the contour line L1 in the vertical direction (x-axis direction), the contour line L2 in the horizontal direction (y-axis direction), the contour line L3 in the oblique direction, and the contour of the curved shape.
  • the line L4 is often composed of at least three pixels, that is, three or more pixels.
  • the comparison pixel includes the pixel to be processed, and selects the pixel at the position that becomes the above-described contour lines L1, L2, and L3.
  • the comparison pixel is set in advance in the pixel space with the processing target pixel as the origin, the x-axis direction, the y-axis direction, the axial direction rotated in the clockwise direction around the origin, and the counterclockwise direction around the origin. It is selected from the pixels in the axial direction rotated by the angle.
  • the comparison pixel group is described as being composed of nine pixels.
  • the number of pixels in the comparison pixel group requires a minimum of three comparison pixels, and includes the pixel to be processed. It is necessary to set as 4 or more.
  • the noise reduction unit 13 determines that noise is superimposed on a pixel when the comparison pixel group including nine pixels belongs to the upper group or the lower group.
  • the processing target pixel S0 is included in the contour line, if the contour line exists, the processing target pixel S0 is composed of three or more pixels. Since the probability of belonging to the middle group within the third and lower ranks is high, the probability that the noise removal process is performed is also low.
  • the upper group and the lower group are each one, and the middle group is configured by seven pixels. Accordingly, it is possible to further increase the probability that the processing target pixel S0 belongs to the middle group as compared to the case where the above-described upper group and lower group are two.
  • the comparison pixel group is configured by 9 pixels including the pixel to be processed has been described.
  • the number of pixels and the pixels of the upper group and the lower group when the number of pixels is used are described.
  • the number may be set based on the result of noise removal by performing an experiment in advance in a plurality of combinations depending on what environment and what kind of image is transmitted.
  • the number of pixels included in each of the upper group, middle group, and lower group may be set according to the type of video signal.
  • the number of pixels in the upper group and the lower group need not be the same.
  • the number of pixels in the group is determined according to the processing result of noise removal from the combination when determining the number of pixels. Also good.
  • the position of the comparison pixel to be compared with the processing target pixel in the comparison pixel group may be set in the pixel space based on the processing result performed in advance. For this reason, the position of the comparison pixel does not have to be symmetrical with respect to the processing target pixel.
  • the position of the comparison pixel in the comparison pixel group in FIG. 3 is that the coordinates of the processing target pixel S0 are (F0, x2, y2), the coordinates of the pixel S5 are (F0, x1, y2), and the coordinates of the pixel S1 are (F0).
  • the process symmetrical pixel S0 and the pixel S5 are separated by 3 pixels in the x-axis direction (that is, two pixels exist between the process target pixel S0 and the pixel S4), and the process target pixel S0 and the pixel S1 Is one pixel apart in the x-axis direction (that is, pixel 1 and processing target pixel S0 are adjacent), and processing target pixel S0 and pixel S7 are four pixels apart in the y-axis direction (that is, processing target pixel S0 and pixel S7 There are three pixels in between, and the processing target pixel S0 and the pixel S3 are separated by two pixels in the y-axis direction (that is, one pixel exists between the pixel S3 and the processing target pixel S0). is doing.
  • FIG. 5 is a block diagram showing a configuration example of the noise reduction unit 13 in FIG.
  • the comparison pixel group has the number of pixels shown in FIG. 3, and there are three pixels between the processing target pixel S0 and the pixel S5, and there is one pixel between the processing target pixel and the pixel S1.
  • the circuit configuration corresponds to an arrangement in which the target pixel S0 and the pixels S7 and S3 are adjacent to each other.
  • the noise reduction unit 13 is provided for each of the luminance signal Y, the color difference signal Cb, and the color difference signal Cr.
  • numerical values of the luminance signal Y, the color difference signal Cb, and the color difference signal Cr will be described as pixel values for each pixel.
  • the noise reduction unit 13 includes line memories LM1, LM2, shift registers SF1, SF2, SF3, a noise determination unit 131, a noise value detection unit 132, a noise reduction calculation unit 133, and a control unit 134.
  • the control unit 134 determines whether or not the processing of the pixel value of the new frame is started based on the vertical synchronization signal supplied from the Y / C separation unit 12, and the line memories LM1 and LM2, and the shift register Pixel value input processing from SF1 to SF3 is performed.
  • the line memories LM1 and LM2 have a capacity for storing in series the pixel values of the number of liquid crystal elements PIX arranged on one main scanning line. That is, when the first row of the main scanning line is input to the line memory LM1, the second row of the main scanning line is input to the line memory LM2, and each time a pixel value is input, Pixel values are shifted in the W direction.
  • Each of the shift registers SF1, SF2, and SF3 is a shift register in which registers each having an 8-bit width (because the gradation is 256 gradations) are connected in seven stages in series, and are processed pixels in the comparison pixel group shown in FIG.
  • the pixel values of the pixels at the respective arrangement positions of the other pixels S1 to S8 with respect to S0 are configured to be output to a noise determination unit 131 described later.
  • the registers SR11, SR15, SR17 of the shift register SF1 the registers SR21, SR27 of the shift register SF2, and the registers SR31, SR35, SR37 of the shift register SF3, the pixels S6, S7, S8, S5 in FIG. , S1, S4, S3, and S2 are input to the noise determination unit 131.
  • the pixel value of the processing target pixel S0 is input to the noise determination unit 131 from the register SR25 of the shift register SF2.
  • the noise determination unit 131 determines the size of the pixel values of the processing target pixel S0 and the pixels S1 to S7, and determines the number of pixels having a pixel value larger than the processing target pixel S0 and the processing target pixel from the determination result of the size relationship. Count each with the number of pixels less than S0. In addition, the noise determination unit 131 sets two pixels with the highest rank in the array (rank 1 of the maximum value and rank 2 next to the maximum value) as an upper group, and two pixels with the low rank in the array (rank 9 of the lowest value). The pixels in the rank 8) that are one above the lowest value are set as the lower group, and the five pixels from the third position to the seventh position in the array are set as the middle group.
  • the noise determination unit 131 belongs to the upper group, and one pixel having a pixel value smaller than that of the processing target pixel S0. If it is within the range, it is determined that the processing target pixel S0 belongs to the lower group. Then, when the processing target pixel S0 belongs to the upper group or the lower group, the noise determination unit 131 determines that noise is superimposed on the processing target pixel S0. Next, the noise determination unit 131 calculates a mean value of pixel values of pixels belonging to the middle group, subtracts the mean value from the pixel value of the processing target pixel S0, and is a difference value obtained by subtraction.
  • the pixel difference value is output to the noise value detection circuit 132.
  • the noise determination unit 131 indicates a noise flag indicating whether noise is superimposed on the processing target pixel S0 (whether it belongs to either the upper group or the lower group) Is added to the noise reduction calculation unit 133.
  • the noise determination unit 131 adds the noise flag to the pixel value of the register SR25 in the shift register SF2 and sequentially outputs it to the noise reduction calculation unit 133.
  • the noise value detection circuit 132 obtains an absolute value of a pixel difference value corresponding to the processing target pixel S0 that is determined to be superimposed with noise supplied from the noise determination unit 131, and the absolute value of this difference value is set to 1.
  • the noise value is calculated by adding the frames and dividing the addition result by the number of supplied difference values.
  • the noise reduction calculation unit 133 causes the noise value detection circuit 132 to display the previous frame.
  • the noise value calculated in (for example, the immediately preceding frame) is read.
  • the noise reduction calculation unit 133 performs a calculation of subtracting the noise value from the supplied pixel value.
  • the noise reduction calculation unit 133 performs a calculation of adding the noise value to the supplied pixel value.
  • the noise reduction calculation unit 133 outputs the calculation result as a video signal Sout subjected to noise removal processing.
  • the noise reduction calculation unit 133 outputs the pixel value supplied from the noise determination unit 131 as it is as the video signal Sout.
  • FIG. 6 is a flowchart for explaining an operation example of image processing in the display device 1.
  • the detection unit 11 is supplied with the broadcast wave signal received from the antenna and outputs the signal to the Y / C separation unit 12.
  • the Y / C separation unit 12 demodulates the supplied signal, performs Y / C separation, performs A / D conversion, and performs video signal (luminance signal Y, color difference signal Cb, color difference) as digital signals.
  • Signal Cr is output to the noise reduction unit 13 for each pixel value (step S101).
  • the noise reduction unit 13 performs noise removal processing on the video signal supplied from the Y / C separation unit 13 to reduce noise (step S102). Then, the noise reduction unit 13 outputs the video signal from which noise has been removed to the RGB conversion unit 14.
  • the RGB conversion unit 14 is suitable for progressive display of video created for I (Interlace) / P (Progressive) conversion (interlace video device) from the video signal from which noise has been removed. (Converted into video) (step S103). Then, the RGB conversion unit 14 adjusts the number of pixels by scaling processing on the I / P converted video signal, and then converts the video signal into an RGB signal (gradation data for each of red, green, and blue). The data is converted and output to the liquid crystal drive unit 15.
  • the liquid crystal drive unit 15 generates a clock signal for writing the supplied RGB signals to the liquid crystal elements PIX arranged in a matrix in the liquid crystal panel 16 (step S104).
  • the liquid crystal driving unit 15 converts the gradation data in the RGB signal into a gradation voltage for performing liquid crystal driving (step S105).
  • the liquid crystal driver 15 holds the gradation voltage for each source line in the liquid crystal panel 16 by an internal hold circuit.
  • the liquid crystal drive unit 15 supplies a predetermined voltage to one of the gate lines in the liquid crystal panel 16 in synchronization with the generated gate clock signal, and applies a processing voltage to the gate electrode of the TFT of the liquid crystal element ( Step S106).
  • the liquid crystal driving unit 15 supplies the gradation voltage held for each source line in the liquid crystal panel 16 (step S107).
  • the gradation voltage is sequentially supplied to the source line within the time when each gate line is selected, and the gradation voltage (gradation degree data) necessary for display is turned on.
  • Data is written to the pixel element connected to the drain of the TFT in the state.
  • the pixel element changes the transmittance by controlling the orientation of the internal liquid crystal according to the applied gradation voltage.
  • the video signal received by the detector 11 is displayed on the liquid crystal panel 16 (step S108). This flowchart is complete
  • FIG. 7 is a flowchart showing an operation example of the noise reduction process in step S102 of FIG.
  • the control unit 134 determines whether or not the processing of the pixel value of the new frame is started based on the vertical synchronization signal supplied from the Y / C separation unit 12. When it is determined that it is the start, the noise reduction process is started. That is, the control unit 134 performs input processing of pixel values (luminance signal Y and color difference signals Cb and Cr) as the supplied video signal Sin to the line memories LM1 and LM1 and the shift registers SF1 to SF3.
  • pixel values luminance signal Y and color difference signals Cb and Cr
  • the noise reduction process is performed for each of the luminance signal Y and the color difference signals Cb and Cr.
  • the noise reduction process is performed for each of the luminance signal Y and the color difference signals Cb and Cr.
  • the same processing is performed in parallel for the pixel values of the remaining signals not described.
  • the comparison pixels for comparing pixels on the upper and lower main scanning lines adjacent to the main scanning line with the processing target pixel are used.
  • a plurality of main scanning lines are selected from the main scanning line with the processing target pixel.
  • the main scanning line may be set up and down with respect to the main scanning line having the processing target pixel.
  • the pixel to be compared with the processing target pixel does not need to be at the target position with respect to the processing target pixel, and the main scanning line with the pixel to be compared with the main scanning line with the processing target pixel. It does not have to be in the target position.
  • the control unit 134 outputs a control signal that instructs the noise determination unit 131 to determine whether noise is superimposed on the processing target pixel. Then, the noise determination unit 131 includes comparison pixels S6, S7, SR7, SR15, and SR17 of the shift register SF1, registers SR21, SR25, and SR27 of the shift register SF2, and shift registers SR31, SR35, and SR37, respectively.
  • the pixel values of S8, S0, processing target pixel S1, and comparison pixels S4, S2, S3, and S5 are input.
  • the noise determination unit 131 compares the pixel values of the processing target pixel S0 and each of the comparison pixels S1 to S8 in the comparison pixel group, and the number of comparison pixels having a pixel value larger than the processing target pixel S0. And the number of comparison pixels having pixel values smaller than the processing target pixel S0 are counted (step S201).
  • the noise determination unit 131 performs processing based on the number of comparison pixels having a pixel value larger than the processing target pixel S0 and the number of comparison pixels having a pixel value smaller than the processing target pixel S0, which is a count result in the comparison pixel group. It is determined whether each of the target pixel S0 and the comparison pixels S1 to S8 belongs to an upper group, a middle group, or a lower group (step S202). At this time, the noise determination unit 131 sets the pixels having the pixel values of the first and second ranks from the top as the upper group, and sets the pixels having the pixel values of the ninth and eighth ranks from the lower as the lower group. The pixels having the pixel values from the third position to the seventh position are set as a middle group.
  • the noise determination unit 131 compares the preset number of pixels as the upper group with the number of comparison pixels having a pixel value larger than the processing target pixel S0, and compares the pixels having a pixel value larger than the processing target pixel S0. Is smaller than a preset number of pixels for the upper group, it is determined that the processing target pixel S0 belongs to the upper group. On the other hand, the noise determination unit 131 determines that the processing target pixel S0 does not belong to the upper group if the number of comparison pixels having a pixel value larger than the processing target pixel S0 is larger than a preset number of pixels to be the upper group. To do.
  • the noise determination unit 131 compares the number of pixels set as a lower group set in advance with the number of comparison pixels having a pixel value smaller than the processing target pixel S0, and the comparison pixel having a pixel value smaller than the processing target pixel S0. Is smaller than the number of pixels to be set as a lower group in advance, it is determined that the processing target pixel S0 belongs to the lower group. On the other hand, the noise determination unit 131 determines that the processing target pixel S0 does not belong to the lower group if the number of comparison pixels having a pixel value smaller than the processing target pixel S0 is larger than the number of pixels set as a lower group set in advance. To do. Then, when the processing target pixel S0 does not belong to either the upper group or the lower group, the noise determination unit 131 determines that the processing target pixel S0 belongs to the middle group.
  • the noise determination unit 131 determines whether or not the processing target pixel S0 belongs to either the upper group or the lower group (step S203). At this time, if the processing target pixel S0 belongs to either the upper group or the lower group, the noise determination unit 131 determines that noise is superimposed, and the process proceeds to step S204. On the other hand, when the processing target pixel S0 belongs to the middle group, the noise determination unit 131 determines that noise is not superimposed and advances the process to step S207.
  • the noise determination unit 131 obtains an average value of the pixel values of the middle group (step S204), obtains a pixel difference value between the pixel value and the average value of the processing target pixel S0 (step S205), and this pixel difference The value is output to the noise value detection circuit 132.
  • the noise determination unit 131 processes noise in the pixel value of the processing target pixel S0 output from the register SR25 of the shift register SF2.
  • a noise flag indicating that the pixel is superimposed on the target pixel S0 is added (step S206) and output to the noise reduction calculation unit 133.
  • the noise determination unit 131 outputs a noise flag indicating that noise is not superimposed on the processing target pixel S0 from the register SR25 of the shift register SF2. Is added to the pixel value to be output (step S207).
  • the noise reduction calculation unit 133 performs noise removal processing on the pixel value of the supplied pixel corresponding to the value of the noise flag (step S208). That is, when the noise flag added to the pixel value indicates that the noise is not superimposed on the processing target pixel S0, the noise reduction calculation unit 133 outputs the supplied pixel value as it is as the output video signal Sout. On the other hand, when the noise flag added to the pixel value indicates that the noise is superimposed, the noise reduction calculation unit 133 reads the noise amount obtained in the previous frame from the noise value detection circuit 132, and processes the pixel S0 to be processed. The pixel value is corrected.
  • the noise reduction calculation unit 133 subtracts the noise amount from the pixel value of the processing target pixel S0, while the data indicating the lower group is included in the noise flag. If added, the amount of noise is added to the pixel value of the processing target pixel S0, and the calculated result is output as the output video Sout.
  • the noise value detection circuit 132 obtains the absolute value of the pixel difference value input from the noise determination unit 131, adds the absolute value to obtain the addition absolute value, and the vertical synchronization signal is obtained from the Y / C separation circuit 12.
  • the noise reduction calculation unit 133 uses the noise amount obtained in the immediately previous frame for noise removal of each pixel in the current frame.
  • the noise reduction calculation unit 133 stores a new noise amount obtained from the pixel difference value obtained in the immediately preceding frame in the noise amount register, and further obtains it in the previous frame. Overwrite the amount of noise.
  • the noise determination unit 131 includes the processing target pixel S0 in the upper group, and the pixel value of the processing target pixel S0 is included in the pixel value and the comparison pixel group of the processing target pixel S0.
  • the processing target pixel S0 When the pixel value is the maximum value among the pixel values of all the comparison pixels, the processing target pixel S0 is included in the lower group, and the pixel value of the processing target pixel S0 is included in the comparison pixel group and the pixel value of the processing target pixel S0.
  • the pixel value is the minimum value among the pixel values of all the comparison pixels included, an average value of pixel value differences between the processing target pixel S0 and the other pixels is obtained, and this average value is used as the pixel difference value.
  • You may comprise so that it may output to the value detection circuit 132.
  • FIG. The process of obtaining the noise amount of the noise value detection circuit 132 is the same as that in the above-described embodiment.
  • the noise determination unit 131 sets a plurality of pixels at other positions in the vicinity different from the pixels of the comparison pixel group as a second comparison pixel group in advance, and performs processing.
  • the target pixel S0 is included in the upper group
  • the processing target pixel S0 and the second comparison pixel An average value of pixel value differences with respect to each of the pixels in the group may be obtained, and this average value may be output to the noise value detection circuit 132 as a pixel difference value.
  • the process of obtaining the noise amount of the noise value detection circuit 132 is the same as that in the above-described embodiment.
  • the pixels of the second comparison pixel group for example, the combination of the registers SR12 and SR15 to SR16 in the shift register SF1, the registers SR22 and SR25 to SR26 in the shift register SF2, and the registers SR32 and SR35 to SR36 in the shift register SF3. Is used.
  • the processing target pixel and the pixels around the processing target pixel constitute a comparison pixel group, and the processing target pixel is in the comparison target group. Since noise removal is performed only when it belongs to either a group or a lower group, the probability that noise is superimposed on the processing target pixel corresponding to the contour line is reduced, and the contour line is unnecessary as noise. It is possible to perform noise removal while suppressing deterioration in image quality by reducing the occurrence of image blurring by erasing. That is, according to the present embodiment, the pixel grouping process based on the pixel values described above can detect the contour line of the image in any direction up, down, left, and right, so that the contour line is not unnecessarily erased. Image noise removal processing can be performed.
  • FIG. 8 is a diagram illustrating an image obtained as a result of noise removal performed on the original image of FIG. 22 according to the present embodiment. Comparing the image of FIG. 8 with the image of FIG. 23, the noise in the image of FIG. 8 is reduced with respect to the image of FIG. 23, and the noise removal processing of this embodiment is more effective than the conventional example. It turns out that it is done effectively. Further, when the image of FIG. 8 is compared with the image of FIG. 24, the noise in the image of FIG.
  • FIG. 9 is a diagram illustrating an image obtained as a result of performing noise removal on the original image in FIG. 25 according to the present embodiment.
  • the image of FIG. 9 is compared with the image of FIG. 26, the blur of the image of FIG. 9 is suppressed with respect to the image of FIG. 26, and there is a dent that can be recognized by the difference in color in the region denoted by reference numeral R1 of FIG.
  • R1 reference numeral
  • the liquid crystal display device according to the second embodiment has the same configuration as that shown in FIG. 1 in the first embodiment.
  • the same reference numerals are given to the same configurations as those in the first embodiment, and only configurations and operations different from those in the first embodiment (configurations and operations of the noise reduction unit) will be described.
  • FIG. 10 is a diagram for explaining an outline of noise removal in the present embodiment.
  • H is the horizontal direction
  • V is the vertical method
  • F is the frame direction.
  • inter-frame detection is defined as checking whether the same pixel in the preceding and following frames has a peak and checking whether it is larger than the surrounding pixels in the preceding and following frames.
  • noise removal processing is performed using a combination of a two-dimensional plane of each frame and frames before and after the two-dimensional plane as a pixel space in which pixels are arranged. For example, as shown in FIG. 10, the frame F2 where the processing target pixel S0 is located and the frame F0 where the processing target pixel S0 is located are compared by a combination of pixels located in the immediately preceding frame F1 and the immediately following frame F2.
  • a pixel group is configured.
  • comparison pixels constituting the comparison pixel group in the frame F1, pixels S10, S11, S12, S13, S14, S15, S16, S17, and S18 around the pixel corresponding to the processing target pixel S0 in the frame F0 are set. It is set. In addition, in the frame F2, pixels S20, S21, S22, S23, S24, S25, S26, S27, and S28 around the pixel corresponding to the processing target pixel S0 in the frame F0 are set as the comparison pixels constituting the comparison pixel group. is doing.
  • the pixel S10 in the frame F1 and the pixel S20 in the frame F2 are in the same position in the frame of the display screen.
  • each of the pixels S11 to S18 in the frame F1 and each of the pixels S21 to S28 in the frame F2 are in the same position in the frame of the display screen.
  • the noise reduction unit 13 compares the comparison pixel S0 with the 17 pixels of the comparison target pixel S0 in the frame F0, the comparison pixels S10 to S18 in the frame F1, and the comparison pixels S20 to S28 in the frame F2. It constitutes a group.
  • the noise reduction unit 13 determines the magnitude of the pixel value of the comparison target pixel S0 and the comparison pixel among the 17 pixels in the comparison pixel group, and the pixel from the comparison target pixel S0 is the same as in the first embodiment.
  • the number of comparison pixels having a large value and the number of comparison pixels having a pixel value smaller than that of the comparison target pixel S0 are counted.
  • the noise reduction unit 13 determines each pixel based on the number of comparison pixels having a pixel value larger than the comparison target pixel S0 and the number of comparison pixels having a pixel value smaller than the comparison target pixel S0. Pixels are divided into three groups: an upper group, a middle group, and a lower group.
  • the noise reduction unit 13 determines that noise is superimposed on the processing target pixel S0.
  • the processing target pixel S0 belongs to the middle group, this process is performed. It is determined that no noise is superimposed on the target pixel S0.
  • the top five pixels (first to fifth pixels from the top to the fifth) in the above-described rank array are set as a top group, and the bottom five pixels (highest) in the rank array.
  • the pixels from the 15th to the 19th pixel from the lowest to the fifth) are the lower group, and the remaining 8 pixels (the 6th to 14th pixels) are the middle group.
  • the contour lines of the figure in the image are the contour line L1 in the vertical direction (x-axis direction), the contour line L2 in the horizontal direction (y-axis direction), and the contour line in the oblique direction.
  • L3 and the contour line L4 having a curved shape it is considered that there are many cases where at least three pixels, that is, the total of the processing target pixel and the comparison pixel is composed of three or more pixels. That is, since there is one processing target pixel, it is necessary to set three or more comparison pixels and to set four comparison pixel groups.
  • the positional relationship of the pixels between frames is that the positions of the comparison pixel and the processing target pixel are the same in each frame, so that the contour line is the previous frame.
  • the detection is made from a total of seven, three, one in the frame where the pixel to be processed is located and three in the immediately following frame.
  • the processing target pixel S0 is included in the contour line, that is, when the contour line exists, the processing target pixel S0 is composed of seven or more pixels. Since the probability of belonging to the middle group within the rank is increased, the probability that the noise removal process is performed is also decreased.
  • the noise reduction unit 13 determines that noise is superimposed on the pixel when belonging to the upper group or the lower group. Further, in order to further reduce the probability that noise removal processing is performed on the processing target pixel S0, the upper group and the lower group are each three, and the middle group is configured by 13 pixels. Accordingly, it is possible to further increase the probability that the processing target pixel S0 belongs to the middle group as compared with the case where the above-described upper group and lower group are five.
  • the comparison pixel group has been described as 19 pixels including the processing target pixel.
  • the number of pixels as the upper group and the lower group is as follows. As in the first embodiment, an experiment is performed in advance in a plurality of combinations depending on what kind of image and what kind of image is transmitted, and the result of noise removal is set. Further, the number of pixels in the upper group and the lower group does not need to be the same, and as described above, the number of pixels in the group is set according to the processing result of noise removal from the combination when determining the number of pixels.
  • the position of the comparison pixel in each frame is the same because of the necessity of detecting the contour line, that is, the comparison pixel to be compared. Are the same between frames.
  • the position of the comparison pixel to be compared with the processing target pixel in the comparison pixel group is also set by the above-described processing result performed in advance. For this reason, the position of the comparison pixel does not have to be symmetrical with respect to the processing target pixel.
  • the positions of the other pixels around the pixel S10 corresponding to the comparison pixel group S0 of the frame F0 are 3 pixels apart from the pixel S10 and the pixel S11. Is one pixel apart, pixel S10 and pixel S17 are four pixels apart, and pixel S10 and pixel S13 are two pixels apart.
  • the peripheral pixels selected as the comparison pixels in the pixel space with respect to the pixel S20 corresponding to the processing target pixel S0 are not targets for the pixel S20.
  • FIG. 11 is a block diagram showing a configuration example of the noise reduction unit 13 in FIG. FIG. 11 shows a circuit configuration corresponding to a comparison pixel group constituted by the following 18 pixels and the processing target pixel S0 shown in FIG.
  • the positions of other pixels around the pixel S10 corresponding to the comparison pixel group S0 of the frame F0 are separated by 3 pixels from the pixel S10 and the pixel S15, and the processing target pixel S10 and the pixel S11 are 1 Pixel separation, pixel S10 and pixel S17 are four pixels apart, and pixel S10 and pixel S13 are two pixels apart.
  • the positions of other pixels around the pixel S20 corresponding to the comparison pixel group S0 of the frame F0 are separated by 3 pixels from the pixel S20 and the pixel S25, and the processing target pixel S20 and the pixel S21 are 1 pixel.
  • the pixel S20 and the pixel S27 are separated by 4 pixels, and the pixel S20 and the pixel S23 are separated by 2 pixels.
  • the noise reduction unit 13 is provided for each of the luminance signal Y, the color difference signal Cb, and the color difference signal Cr.
  • numerical values of the luminance signal Y, the color difference signal Cb, and the color difference signal Cr will be described as pixel values for each pixel.
  • the noise reduction unit 13 includes frame memories FM1, FM2, line memories LM11, LM12, LM21, LM31, LM32, shift registers SF11, SF12, SF13, SF21, SF31, SF32, SF33, a noise determination unit 131, and a noise value detection unit 132.
  • a noise reduction calculation unit 133 and a control unit 134 are provided.
  • each frame memory, line memory, and shift register has 8 bits in each of Y, Cr, and Cb because the video has 256 gradations.
  • the control unit 134 determines whether or not the processing of the pixel value of the new frame is started based on the vertical synchronization signal supplied from the Y / C separation unit 12, and the frames FM1 and FM2, the line memory LM11, Pixel value input processing and shift processing are performed in each of the LM12, LM21, LM31, and LM32, and the shift registers SF11, SF12, SF13, SF21, SF31, SF32, and SF33.
  • the frame memories FM1 and FM1 have a capacity for storing pixel values of all the pixels constituting one frame.
  • the line memories LM11 and LM31 have a capacity for storing in series the pixel values of the number of liquid crystal elements PIX arranged on the three main scanning lines.
  • the line memories LM12, LM21, and LM32 have a capacity for five scanning lines.
  • FIG. 11 as for the pixels in the input video signal Sin, the pixel of the previous frame is stored in the frame memory FM1, and the pixel of the immediately previous frame is stored in the frame FM1.
  • Each of the frame memories FM1 and FM2 is formed as a serial shift register to which pixel values of one frame of pixels are sequentially transferred.
  • the pixel values of the pixels input as the video signal Sin are sequentially transferred by the control unit 134 in the shift register that sequentially transfers the pixel values of the pixels for two frames including the frame memories FM2 and FM1. Then, the pixel values transferred from the frame memory FM2 to the frame memory FM1 are sequentially input to the line memory LM21. Further, the pixel values output from the frame memory FM1 are sequentially input to the line memory LM12. Further, in the present embodiment, the processing target pixel S0 from which noise is removed is a pixel having a pixel value of the register SF215 in the shift register SF21.
  • the nth frame is output from the frame memory FM1
  • the (n + 1) th frame is output from the frame memory FM2
  • the noise removal process is started when the n + 2th frame is input as the video signal Sin.
  • the comparison pixel contrast group constituted by the comparison pixel of the frame F1 which is n frame, the comparison pixel of F2 which is n + 2 frame, and the processing target pixel S0 of F0 which is n + 1 frame
  • the shift register SF21 Noise removal processing is performed on the pixel value of the register SF215 (pixel value of the processing target pixel S0).
  • each of the shift registers SF11, SF12, SF13, SF21, SF31, SF32, and SF33 is a shift register in which 8-bit width registers are connected in series in seven stages, and the pixel to be processed in the comparison pixel group shown in FIG.
  • the arrangement positions of the comparison pixels S10 to S18 and S20 to S28 for S0 are set.
  • the registers SF111, SF115, and S117 in the shift register SF11, the registers SF121, SF125, and S127 in the shift register SF12, and the registers SF131, SF135, and S137 in the shift register SF13, respectively are the pixels S16, S17, and S17 in the frame F1, respectively.
  • the pixel values of S18, S15, S10, S11, S14, S13, and S12 are input to the noise determination unit 131. Further, the pixel value of the processing target pixel S0 is input to the noise determination unit 131 from the register SF215 in the shift register SF21.
  • the registers SF311, SF315, and S317 in the shift register SF31, the registers SF321, SF325, and S327 in the shift register SF32, and the registers SF331, SF335, and S337 in the shift register SF33 are the pixels S26, S27, and S28 of the frame F2, respectively.
  • S25, S20, S21, S24, S23, and S22 are input to the noise determination unit 131.
  • the noise determination unit 131 determines the pixel value of the processing target pixel S0 and each of the 19 pixels of the comparison pixels S10 to S18 and S20 to S28, and the number of comparison pixels having a pixel value larger than that of the processing target pixel S0. And the number of comparison pixels having a pixel value smaller than that of the processing target pixel S0 are counted. In addition, the noise determination unit 131 sets the five pixels having the highest rank in the array (the highest rank from the first rank to the fifth rank) as the upper group, and the five pixels having the lower rank in the array (the lowest rank of the 15th rank). 5 pixels in the order 19) are set as a lower group, and 9 pixels from the 6th position to the 14th position in the array are set as a middle group.
  • the noise determination unit 131 sets the five pixels including the processing target pixel S0 as an upper group, while the pixel value of the processing target pixel S0 is higher than that of the processing target pixel S0.
  • the number of comparison pixels having a small value is 4 or less, 5 pixels including the processing target pixel S0 are set as a lower group. Then, when the processing target pixel S0 belongs to either the upper group or the lower group, the noise determination unit 131 determines that noise is superimposed on the processing target pixel S0.
  • the noise determination unit 131 calculates an average value of pixel values in the middle group composed of nine pixels that do not belong to either the upper group or the lower group, and calculates this average from the pixel value of the processing target pixel S0.
  • the pixel difference value obtained by subtracting the value is output to the noise value detection circuit 132.
  • the noise determination unit 131 indicates a noise flag indicating whether noise is superimposed on the processing target pixel S0 (whether it belongs to either the upper group or the lower group) Is added to the noise reduction calculation unit 133.
  • the noise determination unit 131 adds the noise flag to the pixel value of the register SF213 in the shift register SF21 and sequentially outputs it to the noise reduction calculation unit 133.
  • the noise value detection circuit 132 obtains the absolute value of the pixel difference value corresponding to the processing target pixel S0 supplied from the noise determination unit 131 and determined to have noise superimposed thereon, and calculates the absolute value of this pixel difference value.
  • the noise value is calculated by adding one frame and dividing the addition result by the number of supplied pixel difference values.
  • the noise reduction calculation unit 133 causes the noise value detection circuit 132 to perform the entire frame.
  • the noise value calculated in is read.
  • the noise reduction calculation unit 133 performs a calculation to subtract the noise value from the supplied pixel value, and the processing target pixel
  • the noise flag indicates that S0 belongs to the lower group
  • an operation of adding the noise value to the supplied pixel value is performed, and the operation result is output as a video signal Sout subjected to noise removal processing.
  • the noise reduction calculation unit 133 outputs the pixel value supplied from the noise determination unit 131 as it is as the video signal Sout.
  • FIG. 12 shows an operation example of the noise removal processing in the noise reduction unit 13 of the image processing apparatus according to the second embodiment. Since the processing of the entire display device 1 of the image processing apparatus is the same as the flowchart shown in FIG. 6, only the operation of the noise reduction unit 13 different from the first embodiment will be described below.
  • the operation shown in FIG. 12 is started by determining whether or not the processing of the pixel value of the new frame is started by the control unit 134 based on the vertical synchronization signal supplied from the Y / C separation unit 12. .
  • control unit 134 outputs the video signals Sin to the frame memories FM1 and FM2, the line memories LM11, LM12, LM21, LM31 and LM32, and the shift registers SF11, SF12, SF13, SF21, SF31, SF32 and SF33. Performs pixel value input processing.
  • the control unit 134 receives m frames after the power is input or the channel is changed and noise removal is started in the same channel, and in this embodiment, for example, 3 frames or more in this embodiment. Is determined (step S300). At this time, the control unit 134 counts the number of vertical synchronization signals supplied from the Y / C separation unit 12 after the start of noise removal processing in the same channel, and stores the internal synchronization unit as the vertical synchronization signal count number. Is written and stored, and it is determined whether or not the vertical synchronization count number is 3 or more, that is, whether the three frames necessary for the video signal to form the comparison pixel group are input. Determine whether or not.
  • control unit 134 resets the vertical synchronization count number to “0” when the power is input or the channel is changed. Then, when the vertical synchronization count number is 3 or more, the control unit 134 advances the process to step S301, and when the horizontal synchronization count number is less than 3, the control unit 134 advances the process to step S307.
  • the pixels on the upper and lower main scanning lines adjacent to the main scanning line having the pixels S10 and S20 corresponding to the processing target pixels in the frames F1 and F2 are used as the comparison pixels.
  • the main scanning lines from the main scanning line having the pixels S10 and S20 through the plurality of main scanning lines may be set up and down with respect to the main scanning line having the processing target pixel.
  • the pixel to be compared with the processing target pixel does not need to be at the target position with respect to the pixels S10 and S20 at the position corresponding to the processing target pixel, and the main pixel having the processing target pixel is present.
  • the scanning line need not be at the target position with respect to the main scanning line having the pixels S10 and S20 at the position corresponding to the processing target pixel.
  • the control unit 134 outputs a control signal that instructs the noise determination unit 131 to determine whether noise is superimposed on the processing target pixel S0.
  • the noise determination unit 131 includes the registers SF111, SF115, and SF117 of the shift register SF11, the registers SF121, SF125, and SF127 of the shift register SF12, the registers SF131, SF135, and SF137 of the shift register SF13, and the register SF215 of the shift register SF21.
  • the noise determination unit 131 compares the pixel values of the processing target pixel S0 with each of the pixels S10 to S18 and the pixels S20 to S28, and determines the number of comparison pixels having a pixel value larger than that of the processing target pixel S0.
  • the number of comparison pixels having a pixel value smaller than that of the processing target pixel S0 is counted (step S301).
  • the noise determination unit 131 determines whether the pixel values of the arranged processing target pixels S0, pixels S10 to S18, and pixels 20 to 28 belong to any of the upper group, middle group, or lower group. The determination is performed in the same manner as in the first embodiment (step S302).
  • the noise determination unit 131 sets the five pixels having the pixel values of the first to fifth ranks from the top as the upper group and the five pixels having the pixel values of the 19th to 15th ranks from the lower.
  • the lower group is set, and the remaining nine pixels having the pixel values of the sixth to the 14th are set as the middle group.
  • the noise determination unit 131 compares the preset number of pixels in the upper group with five of the number of comparison pixels having a pixel value larger than the processing target pixel S0, and the pixels larger than the processing target pixel S0. If the number of comparison pixels of the value is less than 5 which is a preset number of pixels for the upper group, it is determined that the processing target pixel S0 belongs to the upper group. On the other hand, the noise determination unit 131 determines that the processing target pixel S0 does not belong to the upper group if the number of comparison pixels having a pixel value larger than the processing target pixel S0 is larger than a preset number of pixels to be the upper group. To do.
  • the noise determination unit 131 compares the number of pixels set as a lower group set in advance with the number of comparison pixels having a pixel value smaller than the processing target pixel S0, and the comparison pixel having a pixel value smaller than the processing target pixel S0. If the number of pixels is less than 5 as a preset lower group, it is determined that the processing target pixel S0 belongs to the lower group. On the other hand, if the number of comparison pixels having a pixel value smaller than the processing target pixel S0 is greater than the preset number of pixels of the lower group, the noise determination unit 131 belongs to the lower group. Judge that there is no. Then, when the processing target pixel S0 does not belong to either the upper group or the lower group, the noise determination unit 131 determines that the processing target pixel S0 belongs to the middle group.
  • the noise determination unit 131 determines whether or not the processing target pixel S0 belongs to either the upper group or the lower group (step S303). At this time, if the processing target pixel S0 belongs to either the upper group or the lower group, the noise determination unit 131 determines that noise is superimposed, and the process proceeds to step S304. On the other hand, if the processing target pixel S0 belongs to the middle group, the noise determination unit 131 determines that noise is not superimposed, and advances the process to step S307.
  • the noise determination unit 131 obtains an average value of the pixel values of the middle group pixels (step S304), obtains a pixel difference value between the pixel value and the average value of the processing target pixel S0 (step S305), and The pixel difference value is output to the noise value detection circuit 132.
  • the noise determination unit 131 adds a noise flag indicating that noise is superimposed on the pixel value of the processing target pixel S0 output from the register SF213 of the shift register SF21 (step S306), and performs noise reduction calculation. Output to the unit 133.
  • the noise determination unit 131 determines that the comparison pixel group for the processing target pixel S0 cannot be configured in step S300 (cannot determine whether noise is superimposed), and the processing target pixel S0 is in the middle level. In the case of belonging to a group, a noise flag indicating that noise is not superimposed on the processing target pixel is added to the pixel value output from the register SF215 of the shift register SF21 and output (step S307).
  • the noise reduction calculation unit 133 performs noise removal processing on the pixel value of the supplied pixel corresponding to the value of the noise flag (step S308). That is, when the noise flag added to the pixel value indicates that the noise is not superimposed on the processing target pixel, the noise reduction calculation unit 133 outputs the supplied pixel value as it is as the output video signal Sout. On the other hand, when the noise flag added to the pixel value indicates that the noise is superimposed, the noise reduction calculation unit 133 reads the noise amount obtained in the previous frame from the noise value detection circuit 132, and processes the pixel S0 to be processed. The pixel value is corrected.
  • the noise reduction calculation unit 133 subtracts the noise amount from the pixel value of the processing target pixel S0, while the data indicating the lower group is included in the noise flag. If added, the amount of noise is added to the pixel value of the processing target pixel S0, and the calculated result is output as the output video Sout.
  • the noise value detection circuit 132 obtains the absolute value of the pixel difference value input from the noise determination unit 131, adds the absolute value to obtain the addition absolute value, and the vertical synchronization signal is obtained from the Y / C separation circuit 12.
  • the noise reduction calculation unit 133 uses the noise amount obtained in the immediately preceding frame for noise removal of each pixel in the current frame.
  • the noise reduction calculation unit 133 stores a new noise amount obtained from the pixel difference value obtained in the immediately preceding frame in the noise amount register, and further obtains it in the previous frame. Overwrite the amount of noise.
  • the noise determination unit 131 includes the case where the processing target pixel S0 is included in the upper group and the pixel value is the maximum value in the comparison pixel group, and the processing target pixel S0 is included in the lower group.
  • the average value of the difference between the pixel values of the processing target pixel S0 and the other pixels is obtained, and this average value is output to the noise value detection circuit 132 as the pixel difference value. You may comprise so that it may do.
  • the process of obtaining the noise amount of the noise value detection circuit 132 is the same as that in the above-described embodiment.
  • the noise determination unit 131 sets a plurality of pixels at other positions in the vicinity different from the pixels of the comparison pixel group as a second comparison pixel group in advance, and performs processing.
  • the target pixel S0 is included in the upper group
  • the pixel value of the processing target pixel S0 is the maximum value compared to the second comparison pixel group
  • the processing target pixel S0 is included in the lower group
  • the pixel value is the lowest value
  • an average value of pixel value differences between the processing target pixel S0 and each pixel of the second comparison pixel group is obtained, and this average value is used as the pixel difference value to detect the noise value.
  • It may be configured to output to the circuit 132.
  • the process of obtaining the noise amount of the noise value detection circuit 132 is the same as that in the above-described embodiment.
  • the pixels of the second comparison pixel group include the registers SF112, SF115, and SF116 of the shift register SF11, the registers SF122, SF125, and SF126 of the shift register SF12, the registers SF132, SF135, and SF136 of the shift register SF13, and the shift A combination of pixels of the registers SF312, SF315, and SF316 of the register SF31, the registers SF322, SF325, and SF326 of the shift register SF32, and the registers SF332, SF335, and SF336 of the shift register SF33 is used.
  • the upper group or the lower group each including five pixels including the processing target pixel S0 and including seven pixels constituting the contour line. If the signal belongs to the middle group, the probability that the noise is not superimposed is high. Since the noise removal is executed only when the processing target pixel S0 is in the comparison target group and the pixel value belongs to either the upper group or the lower group, the processing target pixel S0 corresponding to the contour line has noise. It is possible to perform noise removal that suppresses degradation of image quality by reducing the probability of being judged as being superposed and unnecessarily erasing the outline as noise to reduce image blurring. .
  • the pixel grouping process based on the pixel values described above can detect the contour line of the image in any direction up, down, left, and right, so that the contour line is not unnecessarily erased.
  • Image noise removal processing can be performed.
  • FIG. 13 is a diagram illustrating an image obtained as a result of noise removal performed on the original image in FIG. 22 according to the present embodiment. Comparing the image of FIG. 13 with the image of FIG. 24, the noise in the image of FIG. 13 is reduced compared to the image of FIG. 24, and there is less blur compared to the image of FIG. It can be seen that the noise removal processing of the present embodiment is performed more effectively than the conventional example.
  • FIG. 14 is a diagram illustrating an image obtained as a result of performing noise removal on the original image in FIG. 25 according to the present embodiment. Comparing the image of FIG. 14 with the image of FIG. 26, the blur of the image of FIG. 14 is suppressed with respect to the image of FIG. 26, and a dent that can be recognized by the difference in color in the region denoted by reference numeral R1 of FIG. It can be seen that it is clearer than the 26 images.
  • FIG. 15 is a diagram illustrating a configuration example of the noise reduction unit 13 according to the third embodiment.
  • the noise reduction part 13 in FIG. 1 by this embodiment is comprised from 13 A of 1st noise reduction parts, and the 2nd noise reduction part 13B, as shown in FIG.
  • the first noise reduction unit 13A performs the first noise removal process on the pixel value of the pixel in the input video signal Sin, and outputs the result of the noise removal as the video signal So.
  • the second noise reduction unit 13B performs a second noise removal process on the pixel value of the video signal So output from the first noise reduction unit 13A, and outputs the result of the noise removal as a video signal Sout.
  • FIG. 16 is a diagram illustrating a configuration example of the first noise reduction unit 13A.
  • the difference between the first noise reduction unit 13A and the noise reduction unit 13 in the second embodiment is that the amount of noise is obtained from a histogram.
  • the noise determination unit 131A outputs the pixel difference value obtained by the processing described in the second embodiment to the histogram analysis circuit 135A.
  • the noise determination unit 131A is the same as the noise determination unit 131 in the second embodiment in other operations.
  • the control circuit 134A is the same as the control unit 134 of FIG. 11 in the second embodiment.
  • the histogram analysis unit 135A obtains the absolute value of the supplied pixel difference value, determines which range of the plurality of numerical ranges the absolute value of the pixel difference value belongs, and the number of pixel difference values belonging to each range And a histogram is generated for each frame unit. Then, each time a vertical synchronization signal is input from the Y / C separation unit 12, the histogram analysis unit 135A selects a numerical range having the largest number of pixel difference values in the generated histogram, that is, a mode value of pixel difference values. The selected mode value is overwritten and stored in the noise amount register as a new noise amount.
  • the numerical range is divided into a plurality of numerical ranges, for example, from the maximum absolute value of pixel difference values obtained by experimental values to 0, and the lower numerical value in the most frequent numerical range is the noise amount.
  • a numerical value range from the noise amount to the maximum absolute value of the pixel difference value obtained from the experimental value Specifically, even if the maximum value of the absolute value of the pixel difference value obtained by the experiment is 120, if the noise amount is 30, 1 to 30, that is, 1, 2, 3,..., 15,. , 30 are provided as one-difference counters, and 31 or more are combined into one counter. That is, the numerical range of the pixel difference value is 1 by dividing 30 from 1 to 30, and 31 is 31 or more.
  • the noise reduction calculation unit 133A is used for noise removal processing of the processing target pixel S0 using the noise amount obtained in the immediately preceding frame, similarly to the noise reduction calculation unit 133 of the second embodiment.
  • the noise reduction calculation unit 133A is the same as the noise reduction calculation unit 133 in the second embodiment in other operations.
  • FIG. 17 is a diagram illustrating a configuration example of the second noise reduction unit 13B.
  • the difference between the second noise reduction unit 13B and the noise reduction unit 13 in the first embodiment is that the amount of noise is obtained from a histogram.
  • the noise determination unit 131B outputs the pixel difference value obtained by the processing described in the first embodiment to the histogram analysis circuit 135B.
  • the noise determination unit 131B is the same as the noise determination unit 131 in the first embodiment in other operations.
  • the control circuit 134B is the same as the control unit 134 of FIG. 5 in the first embodiment.
  • the histogram analysis unit 135B obtains the absolute value of the supplied pixel difference value, determines which range of the plurality of numerical ranges the absolute value of the pixel difference value belongs, and the number of pixel difference values belonging to each range And a histogram is generated for each frame unit. Then, each time a vertical synchronization signal is input from the Y / C separation unit 12, the histogram analysis unit 135B selects a numerical range with the largest number of pixel difference values in the generated histogram, that is, a mode value of pixel difference values. The noise amount corresponding to this numerical range is stored by overwriting the new noise amount in the noise amount register.
  • the numerical range is the same as the relationship of the noise amount already described in the first noise reduction unit 13A.
  • FIG. 18 is a diagram illustrating an image obtained as a result of performing noise removal on the original image in FIG. 22 according to the present embodiment. Comparing the image of FIG. 18 with the image of FIG. 24, the noise in the image of FIG. 18 is reduced compared to the image of FIG. 24, and there is less blur compared to the image of FIG. It can be seen that the noise removal processing of the present embodiment is performed more effectively than the conventional example.
  • each block of the noise reduction unit 13 may be configured as hardware called a signal processing device by a logic circuit formed on an integrated circuit (IC chip), and the configuration other than the line memory and the frame memory is as follows. As described above, it may be realized by software using a CPU (Central Processing Unit).
  • the display device may be an organic EL display device, a CRT (cathode ray tube) display device, or the like.
  • a program for realizing the functions of the histogram analysis unit 135A, the noise determination unit 131B, the noise reduction calculation unit 133B, the control unit 134B, and the histogram analysis unit 135B in FIG. 17 is recorded on a computer-readable recording medium.
  • Image noise removal processing may be performed by causing a computer system to read and execute a program recorded on a recording medium.
  • the “computer system” includes an OS and hardware such as peripheral devices.
  • the “computer system” includes a homepage providing environment (or display environment) if a WWW system is used.
  • the “computer-readable recording medium” refers to a storage device such as a flexible medium, a magneto-optical disk, a portable medium such as a ROM or a CD-ROM, and a hard disk incorporated in a computer system.
  • the “computer-readable recording medium” dynamically holds a program for a short time like a communication line when transmitting a program via a network such as the Internet or a communication line such as a telephone line.
  • a volatile memory in a computer system serving as a server or a client in that case, and a program that holds a program for a certain period of time are also included.
  • the program may be a program for realizing a part of the functions described above, and may be a program capable of realizing the functions described above in combination with a program already recorded in a computer system.

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Image Processing (AREA)
  • Picture Signal Circuits (AREA)

Abstract

An image processing apparatus comprises: a noise evaluation unit that composes a comparison image group with processing-subject pixels that are to be subjected to noise removal processing in a pixel space formed by a plurality of frames included in a video signal and at least three comparison pixels positioned surrounding the processing-subject pixels in the pixel space, that compares pixel values of a plurality of pixels belonging to the comparison image group, and that evaluates whether the pixel values of the processing-subject pixels belong to an upper group, a medium group, or a lower group; and a noise reduction calculation unit that subtracts a preset amount of noise from the pixel values of the processing-subject pixels when the processing-subject pixels belong to the upper group, adds the amount of noise to the pixel values of the processing-subject pixels when the processing-subject pixels belong to the lower group, and does not add or subtract the amount of noise to/from the pixel values of the processing-subject pixels when the processing-subject pixels belong to the medium group.

Description

画像処理装置、画像処理方法及び表示装置Image processing apparatus, image processing method, and display apparatus
 本発明は、画像処理装置、画像処理方法及び表示装置に関する。
 本願は、2010年11月2日に、日本に出願された特願2010-246676号に基づき優先権を主張し、その内容をここに援用する。
The present invention relates to an image processing device, an image processing method, and a display device.
This application claims priority on November 2, 2010 based on Japanese Patent Application No. 2010-246676 filed in Japan, the contents of which are incorporated herein by reference.
 従来から、画像の撮像や伝送を行う際、いわゆるスノーノイズ、ガウスノイズ、ショットノイズなどの雑音成分が画像信号に混入することが知られている。例えば、アナログテレビジョン信号の放送波を受信する際、受信する放送波の電界強度が弱いと、放送波を受信した受信信号は雑音成分の多い映像信号となる。また、デジタル放送においても、過去に録画したアナログ映像をデジタル化して再放送することも多く、その場合の映像信号は、雑音成分の多い映像信号となる。 Conventionally, it is known that noise components such as so-called snow noise, Gaussian noise, and shot noise are mixed in an image signal when an image is captured and transmitted. For example, when receiving a broadcast wave of an analog television signal, if the electric field strength of the received broadcast wave is weak, the received signal that has received the broadcast wave becomes a video signal with a lot of noise components. Also, in digital broadcasting, analog video recorded in the past is often digitized and rebroadcast, and the video signal in that case is a video signal with a lot of noise components.
 この雑音成分を低減する技術として、ローパスフィルタが知られているが、ローパスフィルタを用いた場合、映像信号における画素の画素値(例えば、輝度信号の値)を平均化処理するため、映像がぼやけてしまうという問題がある。
 すなわち、雑音成分の多い画像においては、雑音低減回路が雑音低減処理を行うことにより、雑音が低減することによって、その点では画質が改善されることになる。しかし、その一方では、エッジがぼやけることによって、その点では画質が劣化してしまう。
 この問題を改善するため、以下に示すメディアンフィルタを用いた適応型ローパスフィルタが提案されている。
As a technique for reducing this noise component, a low-pass filter is known. However, when a low-pass filter is used, the pixel value (for example, the value of a luminance signal) of pixels in the video signal is averaged, so that the video is blurred. There is a problem that it ends up.
That is, in an image having a large amount of noise components, the noise reduction circuit performs noise reduction processing to reduce the noise, thereby improving the image quality. However, on the other hand, the image quality deteriorates at that point due to blurring of the edge.
In order to improve this problem, an adaptive low-pass filter using the median filter shown below has been proposed.
 特許文献1においては、雑音低減回路が映像信号の垂直ブランキング期間における画像信号の無いライン部分の雑音成分を検出して、その雑音成分の電圧(ノイズ量E3)を基にして映像信号における雑音の低減を行うことが示されている。
 具体的には、雑音低減回路は、図19に示すように、遅延回路512からの処理画素の電圧E1と、メディアンフィルタ513においてメディアンフィルタ処理を掛けた後の出力電圧E2とを比較する。
In Patent Document 1, a noise reduction circuit detects a noise component of a line portion having no image signal during a vertical blanking period of the video signal, and noise in the video signal is based on the voltage (noise amount E3) of the noise component. It has been shown to reduce.
Specifically, as shown in FIG. 19, the noise reduction circuit compares the processing pixel voltage E1 from the delay circuit 512 with the output voltage E2 after the median filter processing is performed in the median filter 513.
 処理画素の電圧E1がメディアンフィルタ513の出力電圧E2よりも大きい場合、雑音低減回路は、画像信号の電圧E1からノイズ量E3を減算した電圧を外部に出力する。
一方、処理画素の電圧E1がメディアンフィルタ513の出力電圧E2よりも低い場合、雑音低減回路は、画像信号の電圧E1にノイズ量E3を加算した電圧を外部に出力する。
また、画像信号の電圧とフィルタ後の電圧が等しい場合、雑音低減回路は、処理画素の電圧E1をそのまま外部に出力する。
 図20は、特許文献1における雑音低減回路のノイズ低減処理の結果を説明する図である。同図において、入力波形WINとして、所望信号のエンベローブEに雑音が重畳した雑音波Wが示されている。また、出力波形WOUTとして、雑音波Wのノイズが低減された雑音波Wと、所望信号のエンベローブEが示されている。
 このノイズ低減処理の結果、図20の入力波形WINに示されたノイズが重畳した元の画像信号に比較し、図20の出力波形WOUTに示すように、ノイズのピークが低減された画像信号が得られる。
When the voltage E1 of the processing pixel is larger than the output voltage E2 of the median filter 513, the noise reduction circuit outputs a voltage obtained by subtracting the noise amount E3 from the voltage E1 of the image signal.
On the other hand, when the voltage E1 of the processing pixel is lower than the output voltage E2 of the median filter 513, the noise reduction circuit outputs a voltage obtained by adding the noise amount E3 to the voltage E1 of the image signal.
When the voltage of the image signal is equal to the filtered voltage, the noise reduction circuit outputs the processing pixel voltage E1 to the outside as it is.
FIG. 20 is a diagram for explaining the result of the noise reduction processing of the noise reduction circuit in Patent Document 1. In FIG. In the figure, as the input waveform W IN, noise waves W A is shown that noise is superimposed on envelope E A of the desired signal. Further, as the output waveform W OUT, and a noise wave W noise of the noise wave W A is reduced, there is shown a envelope E B of the desired signal.
The result of this noise reduction processing, compared to the original image signal noise which is indicated by the input waveform W IN of Figure 20 is superimposed, as shown by the output waveform W OUT in FIG. 20, image noise peak is reduced A signal is obtained.
 また、特許文献2には、図21に示す構成のノイズ除去装置601を適応型ローパスフィルタとして用いてノイズ除去の処理を行うことが記載されている。すなわち、減算手段612はノイズ除去後の信号(1フレーム遅延信号)Im1と入力信号Di0とから、1フレーム前の差分Diffを求める。動き・ノイズ検出手段614は、この差分Diffが画像の動きによるものか、あるいはノイズによるものなのかの判定を行い、判定結果を動き度合い信号MDSとして出力する。係数変換手段616は、動き度合い信号MDSを巡回係数Kmに変換する。そして、乗算手段617は、振幅制限手段613が差分Diffの振幅を制限した出力Dfnに対し、巡回係数Kmを乗じてノイズ巡回量Ndとする。加算手段618は、入力信号Di0に対し、ノイズ巡回量Ndを加算することにより、フィルタ後の画像信号として出力信号Do0を出力する。 Further, Patent Document 2 describes that noise removal processing is performed using a noise removal device 601 having the configuration shown in FIG. 21 as an adaptive low-pass filter. That is, the subtracting unit 612 obtains a difference Diff one frame before from the signal (1 frame delayed signal) Im1 after noise removal and the input signal Di0. The motion / noise detection unit 614 determines whether the difference Diff is due to image motion or noise, and outputs the determination result as a motion degree signal MDS. The coefficient conversion means 616 converts the motion degree signal MDS into a cyclic coefficient Km. The multiplying unit 617 multiplies the output Dfn whose amplitude limiting unit 613 limits the amplitude of the difference Diff by a cyclic coefficient Km to obtain a noise cyclic amount Nd. The adding means 618 outputs the output signal Do0 as an image signal after filtering by adding the noise circulation amount Nd to the input signal Di0.
 上述したように、ノイズ除去装置601は、出力信号Do0がノイズ除去後の信号Im1として、1フレーム遅延して処理対象の現在の画像信号のフィルタ処理に用いている。このため、フレーム間における画素の差分を画像の動きかノイズかの区別を行い、動き度合い信号MDSに応じてノイズに対する係数を設定してノイズ除去を行うので、画像の動きをノイズとして除去する処理を低減することができる。 As described above, the noise removal apparatus 601 uses the output signal Do0 as a signal Im1 after noise removal for one-frame delayed filtering of the current image signal to be processed. For this reason, the pixel difference between frames is discriminated as image motion or noise, and noise is removed by setting a coefficient for noise according to the motion degree signal MDS. Therefore, processing for removing image motion as noise is performed. Can be reduced.
特開平07-250264号公報Japanese Patent Laid-Open No. 07-250264 特開2010-11482号公報JP 2010-11482 A
 しかしながら、特許文献1及び特許文献2は、雑音低減処理によって画像の精細感が失われたり、画像のエッジ(輪郭線)がぼやけたりするという欠点が依然存在している。
 図22は、撮像した画像データに対して離散的なノイズを人為的に付加した原画像を示す図である。また、図23は特許文献1の雑音低減回路を用い、図22の画像に対してノイズ除去の処理を行った結果を示す図である。この図23に示す画像は、図22に示す原画像に対して、ノイズ除去の処理を横方向(X軸に平行な方向)に1回行ったものであるが、離散的なノイズが女性の顔部分に多く残っており、ノイズ除去が不十分であることが判る。
However, Patent Document 1 and Patent Document 2 still have drawbacks in that the fineness of the image is lost due to the noise reduction processing and the edges (contour lines) of the image are blurred.
FIG. 22 is a diagram illustrating an original image obtained by artificially adding discrete noise to captured image data. FIG. 23 is a diagram showing a result of performing noise removal processing on the image of FIG. 22 using the noise reduction circuit of Patent Document 1. The image shown in FIG. 23 is obtained by performing noise removal processing once in the horizontal direction (direction parallel to the X axis) on the original image shown in FIG. Many remain on the face, indicating that noise removal is insufficient.
 次に、図24は、特許文献1の雑音低減回路を用い、図23の画像に対して縦方向(Y軸に平行な方向)にノイズ処理を行った結果を示す図である。図から判るように、図24の画像は、図23の画像と比較して女性の顔の部分の離散的なノイズが除去され、ノイズ除去の処理が十分に行われている。
 しかしながら、図22の画像と図24の画像とを比較してみると、離散的なノイズが十分除去されているが、一方、図22の画像の輪郭に対して、図24の画像における女性の顔に架かる髪の毛や、左の植物の葉っぱの輪郭がかなりぼけていることが判る。
Next, FIG. 24 is a diagram illustrating a result of performing noise processing in the vertical direction (direction parallel to the Y axis) on the image of FIG. 23 using the noise reduction circuit of Patent Document 1. As can be seen from the figure, the image of FIG. 24 has the noise removal process sufficiently performed by removing the discrete noise of the female face portion as compared with the image of FIG.
However, when comparing the image of FIG. 22 with the image of FIG. 24, the discrete noise is sufficiently removed. On the other hand, the image of FIG. It can be seen that the hair on the face and the outline of the leaf of the plant on the left are quite blurred.
 また、図25は、ITE(The Institute of Image Information and Television Engineers:(社)映像情報メディア学会)標準動画像におけるNo.32の「苔と石仏」の画像の一部を示している。図26は、図25の画像に対して、特許文献1における雑音低減処理を行って得られた画像を示している。図25の画像と図26の画像とを比較してみると、図25の画像では符号R1を付した領域内で色の違いにより認識できる窪みが、図26の画像においてはぼやけて消えかかっていることが判る。
 上述した図22の画像や図25の画像にあるノイズを消去させようとすると、図24の画像または図26の画像に見られるように、ぼけの程度が増加して輪郭線が消えてしまう状態となる。
Further, FIG. 25 shows No. in the standard moving image of ITE (The Institute of Image Information and Television Engineers). A part of the image of 32 “moss and stone Buddha” is shown. FIG. 26 shows an image obtained by performing noise reduction processing in Patent Document 1 on the image of FIG. When comparing the image of FIG. 25 with the image of FIG. 26, in the image of FIG. 25, the dent that can be recognized due to the difference in color within the region denoted by reference numeral R1 is blurred and disappears in the image of FIG. I know that.
When the noise in the image of FIG. 22 or the image of FIG. 25 is deleted, the degree of blur increases and the outline disappears as seen in the image of FIG. 24 or the image of FIG. It becomes.
 そこで、本発明は、上記問題に顧みて成されたものであり、画像における輪郭線の消去を低減させ、惚けを抑制したノイズ除去が行える画像処理装置、画像処理方法及びこの画像処理装置を用いた表示装置を提供することである。 Therefore, the present invention has been made in view of the above problems, and an image processing apparatus, an image processing method, and an image processing apparatus that can eliminate noise with reduced blurring and reduced blur are used. It is to provide a display device.
 この発明は上述した課題を解決するためになされたもので、本発明の一態様による画像処理装置は、映像信号に含まれる複数フレームの画素から形成される画素空間におけるノイズ除去の処理の対象となる処理対象画素と、前記画素空間において当該処理対象画素の周辺に位置する少なくとも3画素の比較画素とから比較画素群を構成し、当該比較画素群に属する複数の画素の画素値を比較し、前記処理対象画素の画素値が上位グループ、中位グループ及び下位グループのいずれに属するかの判定を行うノイズ判定部と、前記処理対象画素が前記上位グループに属する場合、当該処理対象画素の画素値から予め設定されるノイズ量を減算し、前記処理対象画素が前記下位グループに属する場合、当該処理対象画素の画素値から前記ノイズ量を加算し、前記処理対象画素が前記中位グループに属する場合、当該処理対象画素の画素値に対して前記ノイズ量を加減算しないノイズリダクション演算部と、を有することを特徴とする。 The present invention has been made to solve the above-described problems, and an image processing apparatus according to an aspect of the present invention is an object of noise removal processing in a pixel space formed from pixels of a plurality of frames included in a video signal. Forming a comparison pixel group from the processing target pixel and at least three comparison pixels located around the processing target pixel in the pixel space, and comparing pixel values of a plurality of pixels belonging to the comparison pixel group, A noise determination unit that determines whether the pixel value of the processing target pixel belongs to an upper group, a middle group, or a lower group; and when the processing target pixel belongs to the upper group, the pixel value of the processing target pixel When a preset noise amount is subtracted from the pixel, and the processing target pixel belongs to the lower group, the noise amount is calculated from the pixel value of the processing target pixel. Summed, if the processing target pixel belongs to the middle group, and having a noise reduction operation unit, not subtracting the amount of noise to the pixel value of the processing pixel.
 本発明の一態様による画像処理装置は、前記上位グループ、中位グループまたは下位グループそれぞれに含まれる画素数は、映像信号の種類に応じて設定されており、前記ノイズ判定部が前記比較画素群における少なくとも三つの画素の画素値を大きい順に順位付けし、前記上位グループ、前記中位グループ及び前記下位グループの各々に設定された前記画素数に応じて振り分けることを特徴とする。 In the image processing apparatus according to an aspect of the present invention, the number of pixels included in each of the upper group, the middle group, and the lower group is set according to a type of video signal, and the noise determination unit is configured to compare the comparison pixel group. The pixel values of at least three pixels are ranked in descending order and are distributed according to the number of pixels set in each of the upper group, the middle group, and the lower group.
 本発明の一態様による画像処理装置は、前記映像信号の各フレームにおいて、前記処理対象画素と該処理対象画素の周辺の画素とを接続する線が、該フレームに含まれる輪郭線の一部を構成する場合における前記処理対象画素の周辺の画素を、前記比較画素として選択したことを特徴とする。 In the image processing device according to one aspect of the present invention, in each frame of the video signal, a line connecting the processing target pixel and a pixel around the processing target pixel is a part of a contour line included in the frame. A pixel around the processing target pixel in the configuration is selected as the comparison pixel.
 本発明の一態様による画像処理装置は、前記比較画素が、前記処理対象画素に対し、横方向、縦方向及び斜め方向において予め設定された位置の前記画素から選択されることを特徴とする。 The image processing apparatus according to an aspect of the present invention is characterized in that the comparison pixel is selected from the pixels set in advance in a horizontal direction, a vertical direction, and an oblique direction with respect to the processing target pixel.
 本発明の一態様による画像処理装置は、前記比較画素が前記画素空間において前記処理対象画素を原点とし、x軸方向、y軸方向、前記原点を中心として時計方向に予め設定した角度回転させた軸方向、前記原点を中心として反時計方向に予め設定した角度回転させた軸方向にある画素から選択されていることを特徴とする。 In the image processing device according to an aspect of the present invention, the comparison pixel is rotated at a preset angle clockwise in the x-axis direction, the y-axis direction, and the origin in the clockwise direction with the processing target pixel in the pixel space as the origin. The pixel is selected from the pixels in the axial direction, which is rotated in the axial direction by a preset angle counterclockwise around the origin.
 本発明の一態様による画像処理装置は、前記画素空間が時間方向を加えた3次元空間の画素配列を含むことを特徴とする。 An image processing apparatus according to an aspect of the present invention is characterized in that the pixel space includes a pixel array in a three-dimensional space with a time direction added.
 本発明の一態様による画像処理装置は、前記画素空間が時間方向を加えた3次元空間の画素配列を含む場合において、1次のノイズ処理を行い、前記画素空間が前記処理対象画素の存在する2次元空間の画素配列の場合において、2次のノイズ処理を行うことを特徴とする。 An image processing apparatus according to an aspect of the present invention performs first-order noise processing when the pixel space includes a pixel array in a three-dimensional space with a time direction added, and the pixel space includes the processing target pixel. In the case of a pixel array in a two-dimensional space, second-order noise processing is performed.
 本発明の一態様による画像処理装置は、前記予め設定されたノイズ量を算出するノイズ値検出回路をさらに有し、前記ノイズ値検出回路が、前記処理対象画素と、前記画素空間における当該処理対象画素の周辺における複数の他の画素との差分に基づいて、前記ノイズ量を算出することを特徴とする。 The image processing apparatus according to an aspect of the present invention further includes a noise value detection circuit that calculates the preset amount of noise, and the noise value detection circuit includes the processing target pixel and the processing target in the pixel space. The noise amount is calculated based on a difference from a plurality of other pixels around the pixel.
 本発明の一態様による画像処理装置は、前記ノイズリダクション演算部が、前記中位グループの画素の画素値の平均値である中位グループ平均値を求め、前記処理対象画素が前記上位グループまたは前記下位グループに属する場合、当該処理対象画素の画素値から前記中位グループ平均値を減算し、前記ノイズ値検出回路は、前記ノイズリダクション演算部が減算した結果の絶対値を求め、前記2次元配列において選択された前記処理対象画素毎に得られた絶対値を前記2次元配列において平均化することにより前記ノイズ量を求めることを特徴とする。 In the image processing device according to an aspect of the present invention, the noise reduction calculation unit obtains a middle group average value that is an average value of pixel values of the middle group pixels, and the processing target pixel is the upper group or the When belonging to a lower group, the middle group average value is subtracted from the pixel value of the processing target pixel, and the noise value detection circuit obtains an absolute value as a result of subtraction by the noise reduction calculation unit, and the two-dimensional array The noise amount is obtained by averaging the absolute value obtained for each of the processing target pixels selected in step 2 in the two-dimensional array.
 本発明の一態様による画像処理装置は、前記ノイズリダクション演算部が、前記処理対象画素が前記上位グループまたは前記下位グループに属する場合、当該処理対象画素の画素値が当該処理対象画素の画素値及び複数の前記比較画素の画素値のうちで最大値あるいは最小値であるか否かを判定し、最大値あるいは最小値である場合、当該処理対象画素の画素値と、各比較画素との差分の平均値である差分平均値を算出し、前記ノイズ検出部へ出力し、前記ノイズ値検出回路は、前記ノイズリダクション演算部が算出した差分平均値の絶対値を算出し、該算出した前記差分平均値の絶対値を前記画素空間において平均化することにより前記ノイズ量を求めることを特徴とする。 In the image processing device according to an aspect of the present invention, when the processing target pixel belongs to the upper group or the lower group, the noise reduction calculation unit determines that the pixel value of the processing target pixel is the pixel value of the processing target pixel and It is determined whether or not the pixel value of the plurality of comparison pixels is the maximum value or the minimum value. If the value is the maximum value or the minimum value, the difference between the pixel value of the processing target pixel and each comparison pixel A difference average value which is an average value is calculated and output to the noise detection unit, and the noise value detection circuit calculates an absolute value of the difference average value calculated by the noise reduction calculation unit, and the calculated difference average The noise amount is obtained by averaging absolute values of the values in the pixel space.
 本発明の一態様による画像処理装置は、前記ノイズリダクション演算部が、前記処理対象画素が前記上位グループまたは前記下位グループに属する場合、当該処理対象画素の画素値から前記中位グループ平均値を減算することにより差分平均値を算出し、前記ノイズ値検出回路は、前記ノイズリダクション演算部が算出した差分平均値の絶対値を求め、前記差分平均値の絶対値を前記画素空間において平均化することにより前記ノイズ量を求めることを特徴とする。 In the image processing device according to one aspect of the present invention, when the processing target pixel belongs to the upper group or the lower group, the noise reduction calculation unit subtracts the middle group average value from the pixel value of the processing target pixel. The noise value detection circuit calculates an absolute value of the difference average value calculated by the noise reduction calculation unit, and averages the absolute value of the difference average value in the pixel space. The amount of noise is obtained by the following.
 本発明の一態様による画像処理装置は、前記ノイズリダクション演算部が、前記処理対象画素が前記上位グループまたは前記下位グループに属する場合、当該処理対象画素の画素値が最大値あるいは最小値であるか否かを判定し、最大値あるいは最小値である場合、当該処理対象画素の画素値と、前記比較画素との差分の平均値である差分平均値を算出し、前記ノイズ値検出回路は、前記ノイズリダクション演算部が算出した差分平均値の絶対値を求め、1フレーム単位に当該差分平均値の数値毎にヒストグラムを生成し、該生成したヒストグラムの最頻値を前記ノイズ量とすることを特徴とする。 In the image processing device according to an aspect of the present invention, the noise reduction calculation unit may determine whether the pixel value of the processing target pixel is the maximum value or the minimum value when the processing target pixel belongs to the upper group or the lower group. If it is a maximum value or a minimum value, a difference average value that is an average value of differences between the pixel value of the processing target pixel and the comparison pixel is calculated, and the noise value detection circuit The absolute value of the difference average value calculated by the noise reduction calculation unit is obtained, a histogram is generated for each numerical value of the difference average value for each frame, and the mode value of the generated histogram is used as the noise amount. And
 本発明の一態様による画像処理方法は、ノイズ判定部が、映像信号に含まれる複数フレームの画素から形成される画素空間におけるノイズ除去の処理の対象となる処理対象画素と、前記画素空間において当該処理対象画素の周辺に位置する少なくとも3画素の比較画素とから比較画素群を構成し、当該比較画素群に属する複数の画素の画素値を比較し、前記処理対象画素の画素値が上位グループ、中位グループ及び下位グループのいずれに属するかの判定を行うノイズ判定過程と、ノイズリダクション演算部が、前記処理対象画素が前記上位グループに属する場合、当該処理対象画素の画素値から予め設定されるノイズ量を減算し、前記処理対象画素が前記下位グループに属する場合、当該処理対象画素の画素値から前記ノイズ量を加算し、前記処理対象画素が前記中位グループに属する場合、当該処理対象画素の画素値に対して前記ノイズ量を加減算しないノイズリダクション演算過程と、を含むことを特徴とする。 In the image processing method according to one aspect of the present invention, the noise determination unit includes a processing target pixel that is a target of noise removal processing in a pixel space formed from pixels of a plurality of frames included in a video signal, and the pixel space includes the processing target pixel. Comparing a comparison pixel group with at least three comparison pixels located around the processing target pixel, comparing pixel values of a plurality of pixels belonging to the comparison pixel group, and the pixel value of the processing target pixel is an upper group, The noise determination process for determining whether it belongs to the middle group or the lower group, and the noise reduction calculation unit are preset from the pixel value of the processing target pixel when the processing target pixel belongs to the upper group. When the noise amount is subtracted and the processing target pixel belongs to the lower group, the noise amount is added from the pixel value of the processing target pixel. If the processing target pixel belongs to the middle group, characterized in that it comprises a noise reduction operation process without subtracting the amount of noise to the pixel value of the target pixel, a.
 本発明の一態様による画像処理プログラムは、コンピュータに、映像信号に含まれる複数フレームの画素から形成される画素空間におけるノイズ除去の処理の対象となる処理対象画素と、前記画素空間において当該処理対象画素の周辺に位置する少なくとも3画素の比較画素とから比較画素群を構成し、当該比較画素群に属する複数の画素の画素値を比較し、前記処理対象画素の画素値が上位グループ、中位グループ及び下位グループのいずれに属するかの判定を行うノイズ判定ステップと、前記処理対象画素が前記上位グループに属する場合、当該処理対象画素の画素値から予め設定されるノイズ量を減算し、前記処理対象画素が前記下位グループに属する場合、当該処理対象画素の画素値から前記ノイズ量を加算し、前記処理対象画素が前記中位グループに属する場合、当該処理対象画素の画素値に対して前記ノイズ量を加減算しないノイズリダクション演算ステップと、を実行させるための画像処理プログラムである。 An image processing program according to an aspect of the present invention provides a computer with a processing target pixel to be subjected to noise removal processing in a pixel space formed from pixels of a plurality of frames included in a video signal, and the processing target in the pixel space. A comparison pixel group is configured from at least three comparison pixels located around the pixel, and pixel values of a plurality of pixels belonging to the comparison pixel group are compared. A noise determination step for determining whether the pixel belongs to a group or a lower group; and when the processing target pixel belongs to the upper group, a noise amount set in advance is subtracted from a pixel value of the processing target pixel, and the processing When the target pixel belongs to the lower group, the noise amount is added from the pixel value of the processing target pixel, and the processing target pixel If it belongs to the middle group is an image processing program for executing a noise reduction operation step without subtracting the amount of noise to the pixel value of the processing pixel.
 本発明の表示装置は、上記画像処理装置のいずれかを有し、映像信号のノイズ除去を行った後に映像表示を行うことを特徴とする。 The display device of the present invention includes any one of the above image processing devices, and performs video display after removing noise from the video signal.
 この発明によれば、画像のノイズ除去において、画像の輪郭線の消去を低減させ、惚けを抑制したノイズ除去が行える。 According to the present invention, in the noise removal of an image, it is possible to reduce noise by reducing the elimination of the outline of the image and suppressing blurring.
本発明に第1~第3の実施形態における液晶表示装置の機能ブロック図である。FIG. 3 is a functional block diagram of a liquid crystal display device according to first to third embodiments of the present invention. 図1における液晶駆動部15と液晶パネル16との信号の接続関係を示す図である。It is a figure which shows the connection relation of the signal of the liquid crystal drive part 15 and the liquid crystal panel 16 in FIG. 本実施形態におけるノイズ除去の概要を説明するための図である。It is a figure for demonstrating the outline | summary of the noise removal in this embodiment. この上位グループを2個、下位グループを2個とした理由を説明する図である。It is a figure explaining the reason which made two this upper group and two lower groups. 図1におけるノイズリダクション部13の第1の実施形態による構成例を示すブロック図である。It is a block diagram which shows the structural example by 1st Embodiment of the noise reduction part 13 in FIG. 表示装置1における画像処理の動作例を説明するフローチャートである。5 is a flowchart for explaining an operation example of image processing in the display device 1. 図6のステップS102におけるノイズリダクション処理の動作例を示すフローチャートである。It is a flowchart which shows the operation example of the noise reduction process in step S102 of FIG. 図22の原画像に対して第1の実施形態によるノイズ除去を行った結果の図である。It is a figure of the result of having performed the noise removal by 1st Embodiment with respect to the original image of FIG. 図25の原画像に対して第1の実施形態によるノイズ除去を行った結果の図である。It is a figure of the result of having performed the noise removal by 1st Embodiment with respect to the original image of FIG. 第2の実施形態におけるノイズ除去の概要を説明するための図である。It is a figure for demonstrating the outline | summary of the noise removal in 2nd Embodiment. 図1におけるノイズリダクション部13の第2の実施形態による構成例を示すブロック図である。It is a block diagram which shows the structural example by 2nd Embodiment of the noise reduction part 13 in FIG. 第2の実施形態による画像処理装置のノイズリダクション部13におけるノイズ除去の処理の動作例を示すフローチャートである。It is a flowchart which shows the operation example of the process of the noise removal in the noise reduction part 13 of the image processing apparatus by 2nd Embodiment. 図22の原画像に対して第2の実施形態によるノイズ除去を行った結果の図である。It is a figure of the result of having performed the noise removal by 2nd Embodiment with respect to the original image of FIG. 図25の原画像に対して第2の実施形態によるノイズ除去を行った結果の図である。It is a figure of the result of having performed the noise removal by 2nd Embodiment with respect to the original image of FIG. 第3の実施形態によるノイズリダクション部13の構成例を示す図である。It is a figure which shows the structural example of the noise reduction part 13 by 3rd Embodiment. 第3の実施形態における第1ノイズリダクション部13Aの構成例を示す図である。It is a figure which shows the structural example of 13 A of 1st noise reduction parts in 3rd Embodiment. 第3の実施形態における第2ノイズリダクション部13Bの構成例を示す図である。It is a figure which shows the structural example of the 2nd noise reduction part 13B in 3rd Embodiment. 図22の原画像に対して本実施形態によるノイズ除去を行った結果の図である。It is a figure of the result of having performed the noise removal by this embodiment with respect to the original image of FIG. 特許文献1における雑音低減回路の構成を示す図である。It is a figure which shows the structure of the noise reduction circuit in patent document 1. FIG. 特許文献1における雑音低減回路のノイズ低減処理の結果を説明する図である。It is a figure explaining the result of the noise reduction process of the noise reduction circuit in patent document 1. FIG. 特許文献2におけるノイズ除去装置の適応型ローパスフィルタの構成を示す図である。It is a figure which shows the structure of the adaptive low-pass filter of the noise removal apparatus in patent document 2. 撮像した画像データに対して離散的なノイズを人為的に付加した原画像を示す図である。It is a figure which shows the original image which added discrete noise artificially with respect to the imaged image data. 特許文献1の雑音低減回路を用い、図22の画像に対してノイズ除去の処理を行った結果の画像を示す図である。It is a figure which shows the image of the result of having performed the noise removal process with respect to the image of FIG. 22 using the noise reduction circuit of patent document 1. FIG. 特許文献1の雑音低減回路を用い、図22の画像に対して横方向にノイズ除去の処理を行った図23に対し、特許文献1の雑音低減回路を用い、図23に対して縦方向にノイズ処理を行った結果を示す図である。In contrast to FIG. 23 in which noise reduction processing in the horizontal direction is performed on the image in FIG. 22 using the noise reduction circuit in Patent Document 1, the noise reduction circuit in Patent Document 1 is used in the vertical direction with respect to FIG. It is a figure which shows the result of having performed the noise process. ITE標準動画像におけるNo.32の「苔と石仏」の画像の一部を示している。No. in the ITE standard moving image. A part of the image of 32 “moss and stone Buddha” is shown. 図25の画像に対して、特許文献1における雑音低減処理を行って得られた画像を示している。The image obtained by performing the noise reduction process in patent document 1 is shown with respect to the image of FIG.
<第1の実施形態>
 以下、図面を参照して、本発明の第1の実施形態を説明する。図1は、本発明の第1の実施形態における液晶表示装置の機能ブロック図である。
 図1において、液晶表示装置1は、検波部11、Y/C(輝度信号/色信号)分離部12、ノイズリダクション部13、RGB(Red:赤、Green:緑、Blue:青)変換部14、液晶駆動部15及び液晶パネル16を備えている。
<First Embodiment>
Hereinafter, a first embodiment of the present invention will be described with reference to the drawings. FIG. 1 is a functional block diagram of a liquid crystal display device according to the first embodiment of the present invention.
In FIG. 1, the liquid crystal display device 1 includes a detection unit 11, a Y / C (luminance signal / color signal) separation unit 12, a noise reduction unit 13, and an RGB (Red: red, Green: green, Blue: blue) conversion unit 14. The liquid crystal drive unit 15 and the liquid crystal panel 16 are provided.
 検波部11は、一例として、不図示のアンテナから供給される地上アナログテレビジョン放送の複数チャネルの画像データの高周波信号を受け取る。そして、検波部11は、アンテナから供給される高周波信号から希望のチャネルの変調信号を抽出し、抽出した変調信号をベースバンドの信号に変換し、Y/C分離部12へ出力する。 As an example, the detection unit 11 receives high-frequency signals of image data of a plurality of channels of terrestrial analog television broadcast supplied from an antenna (not shown). Then, the detection unit 11 extracts a modulation signal of a desired channel from the high-frequency signal supplied from the antenna, converts the extracted modulation signal into a baseband signal, and outputs it to the Y / C separation unit 12.
 Y/C分離部12は、供給されたベースバンドの信号を復調し、輝度信号Yと、色差信号Cbと色差信号Crとに分離し、分離したそれぞれの信号を、所定のサンプリング周波数でデジタル信号へ変換する。
 また、Y/C分離部12は、デジタル信号に変換した輝度信号Yと、色差信号Cbと色差信号Crとを、画像の主走査方向(横方向、水平方向)に隣接して並ぶ画素毎に同期し、ノイズリダクション部13へ出力する。これら輝度信号Yと、色差信号Cbと色差信号Crとの各々は、電圧値で表現した信号強度を示すデジタル信号による示された数値である。
The Y / C separation unit 12 demodulates the supplied baseband signal, separates it into a luminance signal Y, a color difference signal Cb, and a color difference signal Cr, and converts each separated signal into a digital signal at a predetermined sampling frequency. Convert to
The Y / C separation unit 12 also arranges the luminance signal Y converted into a digital signal, the color difference signal Cb, and the color difference signal Cr for each pixel arranged adjacent to each other in the main scanning direction (horizontal direction, horizontal direction) of the image. Synchronize and output to the noise reduction unit 13. Each of the luminance signal Y, the color difference signal Cb, and the color difference signal Cr is a numerical value indicated by a digital signal indicating a signal intensity expressed by a voltage value.
 ノイズリダクション部13は、供給される輝度信号Y、色差信号Cb、色差信号Crそれぞれを、同一フレーム(画素が配列された画素空間)内の画素間において比較し、処理対象画素にノイズが重畳しているか否かの判定を行う。
 また、ノイズリダクション部13は、フレーム単位あるいはフィールド単位でノイズ量の算出を行う。ノイズリダクション部13は、ノイズが重畳していると判定された処理対象画素から算出したノイズ量を加減算することで、処理対象画素のノイズ除去の処理を行い、処理後の輝度信号Y、色差信号Cb、色差信号Crそれぞれを、同期させ、映像信号としてRGB変換部14へ出力する。
 このノイズリダクション部13における各画素に対するノイズ処理の詳細については後述する。ここで、映像信号がインターレスである場合、フィールド毎にノイズ処理が行われており、一方、映像信号がノンインターレスである場合、フレーム毎にノイズ処理が行われている。
The noise reduction unit 13 compares the supplied luminance signal Y, color difference signal Cb, and color difference signal Cr between pixels in the same frame (pixel space in which pixels are arranged), and noise is superimposed on the processing target pixel. It is determined whether or not.
The noise reduction unit 13 calculates the amount of noise in frame units or field units. The noise reduction unit 13 performs noise removal processing on the processing target pixel by adding and subtracting the noise amount calculated from the processing target pixel determined to have noise superimposed thereon, and the processed luminance signal Y and color difference signal Cb and the color difference signal Cr are synchronized and output to the RGB converter 14 as a video signal.
Details of the noise processing for each pixel in the noise reduction unit 13 will be described later. Here, when the video signal is interlaced, noise processing is performed for each field. On the other hand, when the video signal is non-interlaced, noise processing is performed for each frame.
 RGB変換部14は、供給される映像信号に対し、その信号がインターレース信号であれば、それをプログレッシブ信号に変換する。また、RGB変換部14は、プログレッシブ信号に対し、液晶パネル16の解像度に合わせて、画素数を調整(スケーリング処理)する。
 そして、RGB変換部14は、画素数が調整された映像信号をRGB信号(Red、Green、Blueのカラービデオ信号)に変換し、変換したRGB信号を液晶駆動部15へ出力する。
If the supplied video signal is an interlace signal, the RGB conversion unit 14 converts it into a progressive signal. In addition, the RGB conversion unit 14 adjusts the number of pixels (scaling processing) in accordance with the resolution of the liquid crystal panel 16 with respect to the progressive signal.
Then, the RGB converter 14 converts the video signal with the adjusted number of pixels into an RGB signal (Red, Green, Blue color video signal), and outputs the converted RGB signal to the liquid crystal driver 15.
 液晶駆動部15は、液晶パネル16に供給される映像データを、画面の2次元平面上に表示するためのクロック信号等を生成する。そして、液晶駆動部15は、液晶パネル16に対して生成したクロック信号を供給する。 The liquid crystal drive unit 15 generates a clock signal or the like for displaying video data supplied to the liquid crystal panel 16 on a two-dimensional plane of the screen. Then, the liquid crystal drive unit 15 supplies the generated clock signal to the liquid crystal panel 16.
 次に、図2は、液晶駆動部15と液晶パネル16との信号の接続関係を示す図である。
 この図2に示すように、液晶駆動部15は、ソースドライバ部とゲートドライバ部とを有しており、液晶パネル16において、ソース線19とゲート線18との交差する点に配置された表示素子PIX、すなわちマトリクス上に配列した表示素子PIXの制御を行い、画像表示を行う。この表示素子PIXは、TFT(Thin Film Transistor、薄膜トランジスタ)と、このTFTにより後述する階調度に対応した電圧が書き込まれる液晶の画素素子とから構成されている。
Next, FIG. 2 is a diagram illustrating a signal connection relationship between the liquid crystal driving unit 15 and the liquid crystal panel 16.
As shown in FIG. 2, the liquid crystal driving unit 15 includes a source driver unit and a gate driver unit. In the liquid crystal panel 16, a display arranged at a point where the source line 19 and the gate line 18 intersect with each other. The element PIX, that is, the display element PIX arranged on the matrix is controlled to display an image. The display element PIX includes a TFT (Thin Film Transistor) and a liquid crystal pixel element into which a voltage corresponding to a gradation described later is written by the TFT.
 ソースドライバ部は、供給されたRGB信号(例えば、RGBそれぞれの画素値を示すデジタル信号)を、液晶パネル16のソース線19(列方向の配線)ごとに、内部に設けられたホールド回路で保持する。
 また、ソースドライバ部は、画面の縦方向の配列に対したゲートクロック信号に同期して、保持しているRGB信号から画素素子駆動用に階調化した電圧(ソース信号)を生成し、生成したソース信号を液晶パネル16の液晶素子PIXに接続されているTFTのソース線19に供給する。
 ゲートドライバ部は、液晶パネル16の液晶素子PIXに接続されているTFTのゲート線18(横方向の配線、主走査に対応)を通じて画面の液晶素子PIXの1行分に対して、ゲートクロック信号に同期して、所定のゲート信号を供給する。
The source driver unit holds the supplied RGB signals (for example, digital signals indicating the respective RGB pixel values) for each source line 19 (wiring in the column direction) of the liquid crystal panel 16 by a hold circuit provided therein. To do.
In addition, the source driver unit generates a voltage (source signal) that is grayscaled for pixel element driving from the held RGB signal in synchronization with the gate clock signal for the vertical arrangement of the screen. The source signal thus supplied is supplied to the source line 19 of the TFT connected to the liquid crystal element PIX of the liquid crystal panel 16.
The gate driver unit supplies a gate clock signal to one row of the liquid crystal elements PIX on the screen through TFT gate lines 18 (corresponding to horizontal wiring, main scanning) connected to the liquid crystal elements PIX of the liquid crystal panel 16. A predetermined gate signal is supplied in synchronization with the above.
 液晶パネル16は、アレイ基板と対向基板と液晶とを備える。アレイ基板上のソース線19とゲート線18との交点ごとに、液晶素子PIX、すなわちTFTとTFTのドレイン電極に接続されている画素電極と対向電極(対向基板上のベタ電極により構成されている)からなる画素素子とが1組ずつ配置されている。ここで、画素素子には、画素電極と対向電極との間に液晶が封入されている。また、液晶パネル16は、画素ごとに、すなわち液晶素子PIX3個毎に3原色RGB(Red、Green、Blue)に対応している。そして、液晶パネル16は、そのサブ画素毎に1つずつの前記TFTを有している。 The liquid crystal panel 16 includes an array substrate, a counter substrate, and a liquid crystal. At each intersection of the source line 19 and the gate line 18 on the array substrate, a liquid crystal element PIX, that is, a pixel electrode connected to the TFT and the drain electrode of the TFT, and a counter electrode (consisting of a solid electrode on the counter substrate) ) Are arranged one by one. Here, in the pixel element, liquid crystal is sealed between the pixel electrode and the counter electrode. The liquid crystal panel 16 corresponds to the three primary colors RGB (Red, Green, Blue) for each pixel, that is, for every three liquid crystal elements PIX. The liquid crystal panel 16 has one TFT for each subpixel.
 TFTは、ゲートドライバ部から供給されたゲート信号がゲート電極に供給され、ゲート信号が例えばハイレベルの時、選択されてオン状態となる。TFTのソース電極には、ソースドライバから供給されたソース信号が供給され、TFTがオン状態の場合、TFTのドレイン電極に接続されている画素電極、すなわち画素素子に階調化された電圧が印加される。 The TFT is selected and turned on when the gate signal supplied from the gate driver unit is supplied to the gate electrode and the gate signal is at a high level, for example. The source signal supplied from the source driver is supplied to the source electrode of the TFT, and when the TFT is in an ON state, a gradation voltage is applied to the pixel electrode connected to the drain electrode of the TFT, that is, the pixel element. Is done.
 その階調化された電圧に応じて、画素素子の液晶の配向が変化し、これによって画素素子の領域における液晶の光の透過度が変化する。その階調化された電圧がTFTのドレイン電極に接続されている画素電極と対向電極との間の液晶部分により構成された画素素子の液晶容量に保持され、液晶の配向が維持される。次の信号がソース電極に供給され、階調化された電圧により、維持された電圧値が変更されるまで、液晶の配向が維持されるので、液晶の光の透過度が維持される。 The orientation of the liquid crystal of the pixel element changes in accordance with the gradation voltage, thereby changing the light transmittance of the liquid crystal in the region of the pixel element. The gradation voltage is held in the liquid crystal capacitance of the pixel element formed by the liquid crystal portion between the pixel electrode connected to the drain electrode of the TFT and the counter electrode, and the alignment of the liquid crystal is maintained. The alignment of the liquid crystal is maintained until the next signal is supplied to the source electrode and the maintained voltage value is changed by the gradation voltage, so that the light transmittance of the liquid crystal is maintained.
 以上説明したようにして、液晶パネル16は、供給された映像データを階調表示する。
 なお、ここでは透過型の液晶パネルについて説明したが、これに限らず反射型の液晶パネルを用いてもよい。
As described above, the liquid crystal panel 16 performs gradation display on the supplied video data.
Although the transmissive liquid crystal panel has been described here, the present invention is not limited to this, and a reflective liquid crystal panel may be used.
 次に、図3は、本実施形態におけるノイズ除去の概要を説明するための図である。同図において、Vは垂直方向(Y軸方法)、Hは水平方向(X軸方向)、Fはフレーム方向(時間軸方向)を示している。この図において、(Fi、xj、yk)は、Fiがフレームの時系列の順番、xjがソース線の番号、ykがゲート線(主走査線)の番号であり、Fi、xj及びykは整数を示し、画素空間における画素の配置を示している(後述する図10においても同様)。例えば、処理対象画素S0の存在する(F0,x2,y2)は、処理対象画素S0の存在するフレームがF0フレーム目であり、このF0フレーム目においてx2番目のソース線とy2番目のゲート線との交点にある画素素子PIXの画素の位置を示している。 Next, FIG. 3 is a diagram for explaining an outline of noise removal in the present embodiment. In the figure, V indicates the vertical direction (Y-axis method), H indicates the horizontal direction (X-axis direction), and F indicates the frame direction (time-axis direction). In this figure, (Fi, xj, yk) is Fi in time sequence of frames, xj is a source line number, yk is a gate line (main scanning line) number, and Fi, xj and yk are integers. And the arrangement of pixels in the pixel space (the same applies to FIG. 10 described later). For example, in the case where the processing target pixel S0 exists (F0, x2, y2), the frame in which the processing target pixel S0 exists is the F0 frame, and in this F0 frame, the x2th source line and the y2th gate line The position of the pixel of the pixel element PIX at the intersection is shown.
 本実施形態におけるノイズ除去において、ノイズリダクション部13がノイズ除去を行う際、処理対象の画素、すなわち処理対象画素にノイズが重畳しているか否かの判定を行う処理について説明する。
 ノイズリダクション部13には、液晶パネル16の液晶素子に横方向、すなわち主走査方向(ゲート線18方向)に順番に画素単位で、Y/C分離部12から輝度信号及び色信号が供給される。
 そして、ノイズリダクション部13には、1行目の主走査線に配置された画素全てに対する映像信号の輝度信号及び色信号の入力が完了すると、2行目の主走査線に配置された画素の映像信号が入力され、順次の主走査線に対応する映像信号が入力される。また、ノイズリダクション部13には、1フレーム分の映像信号が供給されると、順次、次のフレームの映像信号が供給される。
In the noise removal in this embodiment, when the noise reduction unit 13 performs noise removal, a process for determining whether noise is superimposed on a pixel to be processed, that is, a pixel to be processed will be described.
The noise reduction unit 13 is supplied with a luminance signal and a color signal from the Y / C separation unit 12 in the horizontal direction, that is, in the main scanning direction (direction of the gate line 18) in units of pixels in order to the liquid crystal elements of the liquid crystal panel 16. .
When the input of the luminance signal and the color signal of the video signal to all the pixels arranged on the first row main scanning line is completed in the noise reduction unit 13, the pixels of the pixels arranged on the second row main scanning line are completed. Video signals are input, and video signals corresponding to sequential main scanning lines are input. Further, when the video signal for one frame is supplied to the noise reduction unit 13, the video signal of the next frame is sequentially supplied.
 本実施形態において、ノイズリダクション部13は、1フレームの2次元平面(2次元空間)に配置された画素間の画素値を比較して、画素にノイズが重畳しているか否かの判定を行う。ノイズリダクション部13は、輝度信号Y、色信号(色差信号Cb、色差信号Cr)の各々に対して独立に、ノイズ除去の処理を行うが、以下、説明を簡単にするため、これらを画素値として説明する。
 例えば、図3において示すように、ノイズリダクション部13は、処理対象画素S0と、処理対象画素S0の周辺の画素S1からS8(比較画素)との9画素により比較画素群を形成する。ノイズリダクション部13は、この比較画素群における9個の画素において、処理対象画素S0と、比較画素S1からS8の各々との画素値の大小を判定し、この大小関係の判定結果から、処理対象画素S0より大きい画素値を有する画素の数と、処理対象画素S0より小さい画素の数との各々を計数する。
In the present embodiment, the noise reduction unit 13 compares pixel values between pixels arranged in a two-dimensional plane (two-dimensional space) of one frame, and determines whether noise is superimposed on the pixels. . The noise reduction unit 13 performs noise removal processing independently on each of the luminance signal Y and the color signal (color difference signal Cb, color difference signal Cr). Will be described.
For example, as illustrated in FIG. 3, the noise reduction unit 13 forms a comparison pixel group by nine pixels including a processing target pixel S0 and pixels S1 to S8 (comparison pixels) around the processing target pixel S0. The noise reduction unit 13 determines the size of the pixel value of the processing target pixel S0 and each of the comparison pixels S1 to S8 in the nine pixels in the comparison pixel group, and determines the processing target from the determination result of the size relationship. Each of the number of pixels having a pixel value larger than the pixel S0 and the number of pixels smaller than the processing target pixel S0 are counted.
 そして、ノイズリダクション部13は、処理対象画素S0より画素値の大きい比較画素の数と、処理対象画素S0より画素値の小さい比較画素の数とを用いて、処理対象画素S0を上位グループ、中位グループ及び下位グループの3つのグループに振り分ける。すなわち、ノイズリダクション部13は、上位グループとする予め設定した画素数より、処理対象画素S0より画素値の大きい比較画素の数が少ない場合、処理対象画素S0を上位グループとする。一方、ノイズリダクション部13は、下位グループとする予め設定した画素数より、処理対象画素S0より画素値の小さい比較画素の数が少ない場合、処理対象画素S0を下位グループとする。また、ノイズリダクション部13は、上位グループとする予め設定した画素数より、処理対象画素S0より画素値の大きい比較画素の数が多く、かつ、下位グループとする予め設定した画素数より、処理対象画素S0より画素値の小さい比較画素の数が多い場合、処理対象画素S0を中位グループとする。
 ノイズリダクション部13は、処理対象画素S0が上位グループまたは下位グループに属した場合、この処理対象画素S0にノイズが重畳されていると判定し、中位グループに属した場合、この処理対象画素S0にノイズが重畳されていないと判定する。
Then, the noise reduction unit 13 uses the number of comparison pixels having a pixel value larger than that of the processing target pixel S0 and the number of comparison pixels having a pixel value smaller than that of the processing target pixel S0, The group is divided into three groups, a rank group and a subordinate group. That is, when the number of comparison pixels having a pixel value larger than that of the processing target pixel S0 is smaller than the preset number of pixels as the upper group, the noise reduction unit 13 sets the processing target pixel S0 as the upper group. On the other hand, the noise reduction unit 13 sets the processing target pixel S0 as the lower group when the number of comparison pixels having a pixel value smaller than that of the processing target pixel S0 is smaller than the preset number of pixels as the lower group. In addition, the noise reduction unit 13 has a larger number of comparison pixels having a pixel value larger than the processing target pixel S0 than the preset number of pixels as the upper group, and a processing target from the preset number of pixels as the lower group. When the number of comparison pixels having a pixel value smaller than that of the pixel S0 is large, the processing target pixel S0 is set as a middle group.
When the processing target pixel S0 belongs to the upper group or the lower group, the noise reduction unit 13 determines that noise is superimposed on the processing target pixel S0. When the processing target pixel S0 belongs to the middle group, the noise reduction unit 13 It is determined that no noise is superimposed on the.
 本実施形態において、ノイズリダクション部13は、上述した順位の配列の上位2個の画素(最上位から2個目までの1位から2位の画素)を上位グループとし、順位の配列の下位2個の画素(最下位から2個目までの9位から8位の画素)を下位グループとし、残りの5画素(3位から7位の画素)を中位グループとする。
 次に、図4は、2次元平面のフレーム内における各画素と、図形の輪郭線との対応関係を示す図である。この図4を用いて、上位グループの画素数及び下位グループの画素数、本実施形態において上位グループを2個、下位グループを2個とした理由を説明する。
 画像における図形の輪郭線は、図4に示すように、垂直方向(x軸方向)の輪郭線L1、水平方向(y軸方向)の輪郭線L2、斜め方向の輪郭線L3及び曲線形状の輪郭線L4などのように、最低3個の、すなわち3個以上の画素から構成される場合が多いと考えられる。このため、本実施形態において、比較画素は、処理対象画素を含んで、上述した輪郭線L1、L2及びL3となる位置の画素を選択する。すなわち、比較画素が画素空間において処理対象画素を原点とし、x軸方向、y軸方向、原点を中心として時計方向に予め設定した角度回転させた軸方向、原点を中心として反時計方向に予め設定した角度回転させた軸方向にある画素から選択されている。ここで、本実施形態においては、比較画素群を9画素で構成されるとして説明するが、比較画素群における画素の数は、最低3個の比較画素を必要とし、また処理対象画素を含めると、4個以上として設定する必要がある。
In the present embodiment, the noise reduction unit 13 sets the upper two pixels (first to second pixels from the top to the second) in the above-described order as an upper group, and the lower two in the order array. The pixels (9th to 8th pixels from the lowest to the second) are set as the lower group, and the remaining 5 pixels (3rd to 7th pixels) are set as the middle group.
Next, FIG. 4 is a diagram showing a correspondence relationship between each pixel in a two-dimensional plane frame and a contour line of a figure. The reason why the number of pixels in the upper group and the number of pixels in the lower group, two upper groups, and two lower groups in this embodiment will be described with reference to FIG.
As shown in FIG. 4, the contour lines of the figure in the image are the contour line L1 in the vertical direction (x-axis direction), the contour line L2 in the horizontal direction (y-axis direction), the contour line L3 in the oblique direction, and the contour of the curved shape. It can be considered that the line L4 is often composed of at least three pixels, that is, three or more pixels. For this reason, in the present embodiment, the comparison pixel includes the pixel to be processed, and selects the pixel at the position that becomes the above-described contour lines L1, L2, and L3. That is, the comparison pixel is set in advance in the pixel space with the processing target pixel as the origin, the x-axis direction, the y-axis direction, the axial direction rotated in the clockwise direction around the origin, and the counterclockwise direction around the origin. It is selected from the pixels in the axial direction rotated by the angle. Here, in the present embodiment, the comparison pixel group is described as being composed of nine pixels. However, the number of pixels in the comparison pixel group requires a minimum of three comparison pixels, and includes the pixel to be processed. It is necessary to set as 4 or more.
 したがって、ノイズリダクション部13は、9個の画素からなる比較画素群において、上位グループあるいは下位グループに属する場合、画素にノイズが重畳していると判定する。
 これにより、輪郭線L1においては、処理対象画素S0が輪郭線に含まれている場合、輪郭線が存在していれば、3個以上の画素で構成されることになり、処理対象画素S0が3位以下で7位以内の中位グループに属する確率が高くなるため、ノイズ除去の処理が行われる確率も低くなる。
 また、輪郭線L1においては、さらに処理対象画素S0に対するノイズ除去の処理が行われる確率を低くする場合、上位グループ及び下位グループをそれぞれ1個とし、中位グループを7個の画素で構成する。このことにより、上述した上位グループ及び下位グループが2個の場合に比較して、処理対象画素S0が中位グループに属する確率をより高くすることもできる。
Therefore, the noise reduction unit 13 determines that noise is superimposed on a pixel when the comparison pixel group including nine pixels belongs to the upper group or the lower group.
Thereby, in the contour line L1, when the processing target pixel S0 is included in the contour line, if the contour line exists, the processing target pixel S0 is composed of three or more pixels. Since the probability of belonging to the middle group within the third and lower ranks is high, the probability that the noise removal process is performed is also low.
In the contour line L1, in order to further reduce the probability of performing noise removal processing on the processing target pixel S0, the upper group and the lower group are each one, and the middle group is configured by seven pixels. Accordingly, it is possible to further increase the probability that the processing target pixel S0 belongs to the middle group as compared to the case where the above-described upper group and lower group are two.
 また、上述の説明においては、比較画素群が処理対象画素を含めて9画素で構成される場合を説明したが、この画素数及びその画素数とした場合の上位グループ及び下位グループとする画素の数を、どのような環境で、どのような画像を送信するかにより、予め複数の組み合わせにて実験を行い、ノイズ除去の処理結果から設定しても良い。例えば、上位グループ、中位グループまたは下位グループそれぞれに含まれる画素数は、映像信号の種類に応じて設定されていても良い。
 また、上位グループと下位グループとの画素の個数は同一の数にする必要がなく、上述したように、グループの画素数の決定の際の組み合わせから、ノイズ除去の処理結果に応じて設定しても良い。
In the above description, the case where the comparison pixel group is configured by 9 pixels including the pixel to be processed has been described. However, the number of pixels and the pixels of the upper group and the lower group when the number of pixels is used are described. The number may be set based on the result of noise removal by performing an experiment in advance in a plurality of combinations depending on what environment and what kind of image is transmitted. For example, the number of pixels included in each of the upper group, middle group, and lower group may be set according to the type of video signal.
Also, the number of pixels in the upper group and the lower group need not be the same. As described above, the number of pixels in the group is determined according to the processing result of noise removal from the combination when determining the number of pixels. Also good.
 また、比較画素群における処理対象画素と比較する比較画素の位置を、上述した予め行う処理結果により画素空間内において設定しても良い。このため、比較画素の位置は、処理対象画素に対して対称の位置にある必要はない。
 例えば、図3における比較画素群における比較画素の位置は、処理対象画素S0の座標が(F0、x2、y2)、画素S5の座標が(F0、x1、y2)、画素S1の座標が(F0、x3、y2)、画素S7の座標が(F0、x2、y1)、画素S3の座標が(F0、x2、y3)である。
 ここで、
 x2-x1=3
 x3-x2=1
 y2-y1=4
 y3-y2=2
 と設定しても良く、すると比較する画素が処理対称画素S0に対して対称に配置されていない。
 すなわち、処理対称画素S0と画素S5とはx軸方向に3画素離れ(すなわち、処理対象画素S0と画素S4との間には2個の画素が存在し)、処理対象画素S0と画素S1とはx軸方向に1画素離れ(すなわち、画素1と処理対象画素S0とは隣接し)、処理対象画素S0と画素S7はy軸方向に4画素離れ(すなわち、処理対象画素S0と画素S7との間には3個の画素が存在し)、処理対象画素S0と画素S3とはy軸方向に2画素離れ(すなわち、画素S3と処理対象画素S0との間には1個の画素が存在し)ている。
Further, the position of the comparison pixel to be compared with the processing target pixel in the comparison pixel group may be set in the pixel space based on the processing result performed in advance. For this reason, the position of the comparison pixel does not have to be symmetrical with respect to the processing target pixel.
For example, the position of the comparison pixel in the comparison pixel group in FIG. 3 is that the coordinates of the processing target pixel S0 are (F0, x2, y2), the coordinates of the pixel S5 are (F0, x1, y2), and the coordinates of the pixel S1 are (F0). , X3, y2), the coordinates of the pixel S7 are (F0, x2, y1), and the coordinates of the pixel S3 are (F0, x2, y3).
here,
x2-x1 = 3
x3-x2 = 1
y2-y1 = 4
y3-y2 = 2
The pixels to be compared are not arranged symmetrically with respect to the processing symmetric pixel S0.
That is, the process symmetrical pixel S0 and the pixel S5 are separated by 3 pixels in the x-axis direction (that is, two pixels exist between the process target pixel S0 and the pixel S4), and the process target pixel S0 and the pixel S1 Is one pixel apart in the x-axis direction (that is, pixel 1 and processing target pixel S0 are adjacent), and processing target pixel S0 and pixel S7 are four pixels apart in the y-axis direction (that is, processing target pixel S0 and pixel S7 There are three pixels in between, and the processing target pixel S0 and the pixel S3 are separated by two pixels in the y-axis direction (that is, one pixel exists between the pixel S3 and the processing target pixel S0). is doing.
 次に、図5を用いて図1におけるノイズリダクション部13の説明を行う。図5は図1におけるノイズリダクション部13の構成例を示すブロック図である。この図5は、比較画素群が図3に示す画素数であり、処理対象画素S0と画素S5との間が3画素であり、処理対象画素と画素S1との間が1画素であり、処理対象画素S0と画素S7及びS3とが隣接している配置に対応した回路構成となっている。
 また、すでに述べたように、このノイズリダクション部13は、輝度信号Yと、色差信号Cbと、色差信号Crとの各々に対して設けられている。以下、これら輝度信号Yと、色差信号Cbと、色差信号Crの信号の数値を、画素毎の画素値として説明する。
Next, the noise reduction unit 13 in FIG. 1 will be described with reference to FIG. FIG. 5 is a block diagram showing a configuration example of the noise reduction unit 13 in FIG. In FIG. 5, the comparison pixel group has the number of pixels shown in FIG. 3, and there are three pixels between the processing target pixel S0 and the pixel S5, and there is one pixel between the processing target pixel and the pixel S1. The circuit configuration corresponds to an arrangement in which the target pixel S0 and the pixels S7 and S3 are adjacent to each other.
Further, as already described, the noise reduction unit 13 is provided for each of the luminance signal Y, the color difference signal Cb, and the color difference signal Cr. Hereinafter, numerical values of the luminance signal Y, the color difference signal Cb, and the color difference signal Cr will be described as pixel values for each pixel.
 ノイズリダクション部13は、ラインメモリLM1、LM2、シフトレジスタSF1、SF2、SF3、ノイズ判定部131、ノイズ値検出部132、ノイズリダクション演算部133及び制御部134とを備えている。 The noise reduction unit 13 includes line memories LM1, LM2, shift registers SF1, SF2, SF3, a noise determination unit 131, a noise value detection unit 132, a noise reduction calculation unit 133, and a control unit 134.
 制御部134は、Y/C分離部12から供給される垂直同期信号により、新たなフレームの画素値の処理が開始されるか否かの判定を行うとともに、ラインメモリLM1及びLM2と、シフトレジスタSF1からSF3とへの画素値の入力処理を行う。
 ラインメモリLM1及びLM2は、1主走査線に配置された液晶素子PIXの数の画素値を直列に記憶する容量を有している。
 すなわち、ラインメモリLM1に主走査線の1行目が入力されると、ラインメモリLM2には主走査線の2行目が入力されることになり、画素値が入力される毎に、順次、W方向に画素値がシフトされていく。
The control unit 134 determines whether or not the processing of the pixel value of the new frame is started based on the vertical synchronization signal supplied from the Y / C separation unit 12, and the line memories LM1 and LM2, and the shift register Pixel value input processing from SF1 to SF3 is performed.
The line memories LM1 and LM2 have a capacity for storing in series the pixel values of the number of liquid crystal elements PIX arranged on one main scanning line.
That is, when the first row of the main scanning line is input to the line memory LM1, the second row of the main scanning line is input to the line memory LM2, and each time a pixel value is input, Pixel values are shifted in the W direction.
 シフトレジスタSF1、SF2及びSF3の各々は、8ビット(階調が256階調のため)幅のレジスタが直列に7段接続されたシフトレジスタであり、図3に示す比較画素群における処理対象画素S0に対する他の画素S1からS8の各々の配置位置の画素の画素値が後述するノイズ判定部131に出力されるように構成されている。
 ここで、シフトレジスタSF1のレジスタSR11、SR15、SR17、シフトレジスタSF2のレジスタSR21、SR27、シフトレジスタSF3のレジスタSR31、SR35、SR37の各々からは、それぞれ図3における画素S6、S7、S8、S5、S1、S4、S3、S2の画素値がノイズ判定部131へ入力される。また、シフトレジスタSF2のレジスタSR25からは、処理対象画素S0の画素値がノイズ判定部131へ入力される。
Each of the shift registers SF1, SF2, and SF3 is a shift register in which registers each having an 8-bit width (because the gradation is 256 gradations) are connected in seven stages in series, and are processed pixels in the comparison pixel group shown in FIG. The pixel values of the pixels at the respective arrangement positions of the other pixels S1 to S8 with respect to S0 are configured to be output to a noise determination unit 131 described later.
Here, from the registers SR11, SR15, SR17 of the shift register SF1, the registers SR21, SR27 of the shift register SF2, and the registers SR31, SR35, SR37 of the shift register SF3, the pixels S6, S7, S8, S5 in FIG. , S1, S4, S3, and S2 are input to the noise determination unit 131. Further, the pixel value of the processing target pixel S0 is input to the noise determination unit 131 from the register SR25 of the shift register SF2.
 ノイズ判定部131は、処理対象画素S0及び画素S1からS7の画素値の大小判定を行い、この大小関係の判定結果から、処理対象画素S0より大きい画素値を有する画素の数と、処理対象画素S0より小さい画素の数との各々を計数する。
 また、ノイズ判定部131は、配列において順位の高い2個(最大値の順位1と最大値の次の順位2)の画素を上位グループとし、配列において順位の低い2個(最低値の順位9と最低値の1つ上の順位8)の画素を下位グループとし、配列における3位から7位の5個の画素を中位グループとする。
The noise determination unit 131 determines the size of the pixel values of the processing target pixel S0 and the pixels S1 to S7, and determines the number of pixels having a pixel value larger than the processing target pixel S0 and the processing target pixel from the determination result of the size relationship. Count each with the number of pixels less than S0.
In addition, the noise determination unit 131 sets two pixels with the highest rank in the array (rank 1 of the maximum value and rank 2 next to the maximum value) as an upper group, and two pixels with the low rank in the array (rank 9 of the lowest value). The pixels in the rank 8) that are one above the lowest value are set as the lower group, and the five pixels from the third position to the seventh position in the array are set as the middle group.
 さらに、ノイズ判定部131は、処理対象画素S0より画素値が大きい画素が1個以内の場合、処理対象画素S0が上位グループに属し、また、処理対象画素S0より画素値が小さい画素が1個以内の場合、処理対象画素S0が下位グループに属していると判定する。そして、ノイズ判定部131は、処理対象画素S0が上位グループあるいは下位グループに属している場合、処理対象画素S0にノイズが重畳されていると判定する。次に、ノイズ判定部131は、中位グループに属する画素の画素値の平均値を算出し、処理対象画素S0の画素値からこの平均値を減算し、減算して得られた差分値である画素差分値をノイズ値検出回路132へ出力する。
 また、ノイズ判定部131は、処理対象画素S0にノイズが重畳されている場合、この処理対象画素S0にノイズが重畳していることを示すノイズフラグ(上位グループまたは下位グループのいずれかに属するかを示す情報も含む)を付加し、ノイズリダクション演算部133へ出力する。
 ここで、ノイズ判定部131は、ノイズリダクション演算部133に対して、シフトレジスタSF2におけるレジスタSR25の画素値に、ノイズフラグを付加して順次出力する。
Furthermore, when the number of pixels having a pixel value larger than that of the processing target pixel S0 is less than one, the noise determination unit 131 belongs to the upper group, and one pixel having a pixel value smaller than that of the processing target pixel S0. If it is within the range, it is determined that the processing target pixel S0 belongs to the lower group. Then, when the processing target pixel S0 belongs to the upper group or the lower group, the noise determination unit 131 determines that noise is superimposed on the processing target pixel S0. Next, the noise determination unit 131 calculates a mean value of pixel values of pixels belonging to the middle group, subtracts the mean value from the pixel value of the processing target pixel S0, and is a difference value obtained by subtraction. The pixel difference value is output to the noise value detection circuit 132.
In addition, when noise is superimposed on the processing target pixel S0, the noise determination unit 131 indicates a noise flag indicating whether noise is superimposed on the processing target pixel S0 (whether it belongs to either the upper group or the lower group) Is added to the noise reduction calculation unit 133.
Here, the noise determination unit 131 adds the noise flag to the pixel value of the register SR25 in the shift register SF2 and sequentially outputs it to the noise reduction calculation unit 133.
 ノイズ値検出回路132は、ノイズ判定部131から供給される、ノイズが重畳していると判定された処理対象画素S0に対応した画素差分値の絶対値を求め、この差分値の絶対値を1フレーム分加算し、供給された差分値の数により加算結果を除算することにより、ノイズ値を算出する。 The noise value detection circuit 132 obtains an absolute value of a pixel difference value corresponding to the processing target pixel S0 that is determined to be superimposed with noise supplied from the noise determination unit 131, and the absolute value of this difference value is set to 1. The noise value is calculated by adding the frames and dividing the addition result by the number of supplied difference values.
 ノイズリダクション演算部133は、シフトレジスタSF2のレジスタSR25から順次供給される処理対象画素S0に、ノイズが重畳していることを示すノイズフラグが付加されている場合、ノイズ値検出回路132が前フレーム(例えば、直前のフレーム)において算出したノイズ値を読み込む。そして、ノイズリダクション演算部133は、処理対象画素S0が上位グループに属していることをノイズフラグが示している場合、供給される画素値からノイズ値を減算する演算を行う。また、ノイズリダクション演算部133は、処理対象画素S0が下位グループに属していることをノイズフラグが示している場合、供給される画素値にノイズ値を加算する演算を行う。そして、ノイズリダクション演算部133は、演算結果をノイズ除去処理を行った映像信号Soutとして出力する。
 一方、ノイズリダクション演算部133は、重畳されていないことを示すノイズフラグが画素値に付加されている場合、ノイズ判定部131から供給される画素値をそのまま、映像信号Soutとして出力する。
When the noise flag indicating that noise is superimposed is added to the processing target pixel S0 sequentially supplied from the register SR25 of the shift register SF2, the noise reduction calculation unit 133 causes the noise value detection circuit 132 to display the previous frame. The noise value calculated in (for example, the immediately preceding frame) is read. Then, when the noise flag indicates that the processing target pixel S0 belongs to the upper group, the noise reduction calculation unit 133 performs a calculation of subtracting the noise value from the supplied pixel value. In addition, when the noise flag indicates that the processing target pixel S0 belongs to the lower group, the noise reduction calculation unit 133 performs a calculation of adding the noise value to the supplied pixel value. Then, the noise reduction calculation unit 133 outputs the calculation result as a video signal Sout subjected to noise removal processing.
On the other hand, when a noise flag indicating that they are not superimposed is added to the pixel value, the noise reduction calculation unit 133 outputs the pixel value supplied from the noise determination unit 131 as it is as the video signal Sout.
 次に、図6に示すフローチャートを用い、表示装置1全体の動作について説明する。図6は、表示装置1における画像処理の動作例を説明するフローチャートである。
 検波部11は、アンテナから受信された放送波の信号が供給され、Y/C分離部12へ出力する。そして、Y/C分離部12は、供給される信号を復調し、Y/C分離を行った後、A/D変換を行い、デジタル信号である映像信号(輝度信号Y、色差信号Cb、色差信号Cr)として、画素値毎にノイズリダクション部13へ出力する(ステップS101)。
Next, the operation of the entire display device 1 will be described using the flowchart shown in FIG. FIG. 6 is a flowchart for explaining an operation example of image processing in the display device 1.
The detection unit 11 is supplied with the broadcast wave signal received from the antenna and outputs the signal to the Y / C separation unit 12. Then, the Y / C separation unit 12 demodulates the supplied signal, performs Y / C separation, performs A / D conversion, and performs video signal (luminance signal Y, color difference signal Cb, color difference) as digital signals. Signal Cr) is output to the noise reduction unit 13 for each pixel value (step S101).
 次に、ノイズリダクション部13は、Y/C分離部13から供給される映像信号のノイズ除去の処理を行い、ノイズを低減する(ステップS102)。
 そして、ノイズリダクション部13は、ノイズが除去された映像信号を、RGB変換部14に対して出力する。
Next, the noise reduction unit 13 performs noise removal processing on the video signal supplied from the Y / C separation unit 13 to reduce noise (step S102).
Then, the noise reduction unit 13 outputs the video signal from which noise has been removed to the RGB conversion unit 14.
 次に、RGB変換部14は、ノイズが除去された映像信号から、I(Interlace)/P(Progressive)変換(インターレース方式の映像装置向けに作成された映像を、プログレッシブ方式での表示に適した映像へと変換)する(ステップS103)。
 そして、RGB変換部14は、I/P変換された映像信号に対してスケーリング処理による画素数の調整を行った後、映像信号をRGB信号(レッド、グリーン及びブルーの各々の階調度データ)に変換して液晶駆動部15へ出力する。
Next, the RGB conversion unit 14 is suitable for progressive display of video created for I (Interlace) / P (Progressive) conversion (interlace video device) from the video signal from which noise has been removed. (Converted into video) (step S103).
Then, the RGB conversion unit 14 adjusts the number of pixels by scaling processing on the I / P converted video signal, and then converts the video signal into an RGB signal (gradation data for each of red, green, and blue). The data is converted and output to the liquid crystal drive unit 15.
 次に、液晶駆動部15は、供給されるRGB信号を、液晶パネル16におけるマトリクス状に配置された液晶素子PIXに書き込むためのクロック信号を生成する(ステップS104)。 Next, the liquid crystal drive unit 15 generates a clock signal for writing the supplied RGB signals to the liquid crystal elements PIX arranged in a matrix in the liquid crystal panel 16 (step S104).
 次に、液晶駆動部15は、RGB信号における階調度データを、液晶駆動を行う階調化された電圧に変換する(ステップS105)。
 そして、液晶駆動部15は、液晶パネル16におけるソース線毎に、その階調化された電圧を、内部のホールド回路により保持する。
Next, the liquid crystal driving unit 15 converts the gradation data in the RGB signal into a gradation voltage for performing liquid crystal driving (step S105).
The liquid crystal driver 15 holds the gradation voltage for each source line in the liquid crystal panel 16 by an internal hold circuit.
 次に、液晶駆動部15は、生成したゲートクロック信号に同期し、所定の電圧を液晶パネル16におけるゲート線のいずれかに供給し、液晶素子のTFTのゲート電極に処理の電圧を印加する(ステップS106)。 Next, the liquid crystal drive unit 15 supplies a predetermined voltage to one of the gate lines in the liquid crystal panel 16 in synchronization with the generated gate clock signal, and applies a processing voltage to the gate electrode of the TFT of the liquid crystal element ( Step S106).
 次に、液晶駆動部15は、生成したゲートクロック信号に同期し、液晶パネル16におけるソース線毎に保持した、階調化された電圧を供給する(ステップS107)。
 上述した処理により、各ゲート線が選択されている時間内に、階調化された電圧がソース線に順次供給され、表示に必要な階調化された電圧(階調度のデータ)が、オン状態にあるTFTのドレインに接続された画素素子に書き込まれる。これにより、画素素子は、印加された階調化された電圧に応じて、内部の液晶の配向が制御されて透過率が変更される。この結果、液晶パネル16には、検波部11が受信した映像信号が表示されることになる(ステップS108)。以上で本フローチャートは終了する。
Next, in synchronization with the generated gate clock signal, the liquid crystal driving unit 15 supplies the gradation voltage held for each source line in the liquid crystal panel 16 (step S107).
By the above-described processing, the gradation voltage is sequentially supplied to the source line within the time when each gate line is selected, and the gradation voltage (gradation degree data) necessary for display is turned on. Data is written to the pixel element connected to the drain of the TFT in the state. As a result, the pixel element changes the transmittance by controlling the orientation of the internal liquid crystal according to the applied gradation voltage. As a result, the video signal received by the detector 11 is displayed on the liquid crystal panel 16 (step S108). This flowchart is complete | finished above.
 次に、図7を用いて、図6におけるノイズリダクション処理について詳細に説明する。
図7は、図6のステップS102におけるノイズリダクション処理の動作例を示すフローチャートである。また、図7のフローチャートに示す動作は、Y/C分離部12から供給される垂直同期信号により、制御部134が新たなフレームの画素値の処理が開始されるか否かの判定を行い、開始であると判定した場合、ノイズリダクション処理が開始される。すなわち、制御部134は、ラインメモリLM1及びLM1と、シフトレジスタSF1からSF3とに対し、供給される映像信号Sinとしての画素値(輝度信号Y及び色差信号Cb、Cr)の入力処理を行う。
 ノイズリダクション処理としては、輝度信号Y及び色差信号Cb、Crの各々について行う。以下、輝度信号Y及び色差信号Cb、Crのいずれか一つの信号の処理を行うとして説明するが、説明しない残りの信号の画素値も同様の処理が並列して行われる。
Next, the noise reduction process in FIG. 6 will be described in detail with reference to FIG.
FIG. 7 is a flowchart showing an operation example of the noise reduction process in step S102 of FIG. In the operation shown in the flowchart of FIG. 7, the control unit 134 determines whether or not the processing of the pixel value of the new frame is started based on the vertical synchronization signal supplied from the Y / C separation unit 12. When it is determined that it is the start, the noise reduction process is started. That is, the control unit 134 performs input processing of pixel values (luminance signal Y and color difference signals Cb and Cr) as the supplied video signal Sin to the line memories LM1 and LM1 and the shift registers SF1 to SF3.
The noise reduction process is performed for each of the luminance signal Y and the color difference signals Cb and Cr. In the following description, it is assumed that any one of the luminance signal Y and the color difference signals Cb and Cr is processed. However, the same processing is performed in parallel for the pixel values of the remaining signals not described.
 本実施形態においては、処理対象画素のある主走査線に隣接する上下の主走査線にある画素を比較する比較画素としたが、処理対象画素のある主走査線から複数本の主走査線を介した主走査線を、処理対象画素のある主走査線に対して上下で設定しても良い。
 また、処理対象画素と比較する画素が、処理対象画素に対して対象の位置にある必要はなく、かつ、処理対象画素のある主走査線に対し、比較する画素のある主走査線に対して対象の位置にある必要はない。比較画素群を構成するために必要な、処理対象画素の存在する主走査線に対し、処理対象画素と比較する画素の主走査線との位置関係及び各主走査線における画素の位置関係を、用いる対象の画像の種類(変化の激しい映像、あまり変化のない映像)などにより、それぞれ実験により設定する。
In the present embodiment, the comparison pixels for comparing pixels on the upper and lower main scanning lines adjacent to the main scanning line with the processing target pixel are used. However, a plurality of main scanning lines are selected from the main scanning line with the processing target pixel. The main scanning line may be set up and down with respect to the main scanning line having the processing target pixel.
Further, the pixel to be compared with the processing target pixel does not need to be at the target position with respect to the processing target pixel, and the main scanning line with the pixel to be compared with the main scanning line with the processing target pixel. It does not have to be in the target position. The positional relationship between the processing target pixel and the main scanning line of the pixel to be processed and the positional relationship of the pixel in each main scanning line with respect to the main scanning line where the processing target pixel exists, which is necessary to configure the comparison pixel group, Depending on the type of target image to be used (video that changes rapidly, video that does not change much), etc., each is set by experiment.
 図7の示すフローチャートにおいて、制御部134は、ノイズ判定部131に対して、処理対象画素にノイズが重畳しているか否かの判定を行うことを指示する制御信号を出力する。
 そして、ノイズ判定部131は、シフトレジスタSF1のレジスタSR11、SR15及びSR17と、シフトレジスタSF2のレジスタSR21、SR25及びSR27と、シフトレジスタSR31、SR35及びSR37との各々から、比較画素S6,S7、S8、S0、処理対象画素S1、比較画素S4、S2、S3、S5の画素値を入力する。
In the flowchart illustrated in FIG. 7, the control unit 134 outputs a control signal that instructs the noise determination unit 131 to determine whether noise is superimposed on the processing target pixel.
Then, the noise determination unit 131 includes comparison pixels S6, S7, SR7, SR15, and SR17 of the shift register SF1, registers SR21, SR25, and SR27 of the shift register SF2, and shift registers SR31, SR35, and SR37, respectively. The pixel values of S8, S0, processing target pixel S1, and comparison pixels S4, S2, S3, and S5 are input.
 次に、ノイズ判定部131は、比較画素群において、処理対象画素S0と、比較画素S1から比較画素S8の各々との画素値を比較し、処理対象画素S0より大きい画素値の比較画素の個数と、処理対象画素S0より小さい画素値の比較画素の個数とを、各々カウントする(ステップS201)。 Next, the noise determination unit 131 compares the pixel values of the processing target pixel S0 and each of the comparison pixels S1 to S8 in the comparison pixel group, and the number of comparison pixels having a pixel value larger than the processing target pixel S0. And the number of comparison pixels having pixel values smaller than the processing target pixel S0 are counted (step S201).
 そして、ノイズ判定部131は、比較画素群におけるカウント結果である、処理対象画素S0より大きい画素値の比較画素の個数と、処理対象画素S0より小さい画素値の比較画素の個数とに基づき、処理対象画素S0、比較画素S1から比較画素S8の各々が、上位グループ、中位グループまたは下位グループのいずれのグループに属するかの判定を行う(ステップS202)。このとき、ノイズ判定部131は、上からの順位が1位及び2位の画素値の画素を上位グループとし、下位から順位が9位及び8位の画素値の画素を下位グループとし、残りの3位から7位の画素値の画素を中位グループとする。すなわち、ノイズ判定部131は、予め設定された上位グループとする画素の個数と、処理対象画素S0より大きい画素値の比較画素の個数とを比較し、処理対象画素S0より大きい画素値の比較画素の数が、予め設定された上位グループとする画素の個数より少なければ、処理対象画素S0が上位グループに属すると判定する。一方、ノイズ判定部131は、処理対象画素S0より大きい画素値の比較画素の数が、予め設定された上位グループとする画素の個数より多ければ、処理対象画素S0が上位グループに属さないと判定する。また、ノイズ判定部131は、予め設定された下位グループとする画素の個数と、処理対象画素S0より小さい画素値の比較画素の個数とを比較し、処理対象画素S0より小さい画素値の比較画素の数が、予め設定された下位グループとする画素の個数より少なければ、処理対象画素S0が下位グループに属すると判定する。一方、ノイズ判定部131は、処理対象画素S0より小さい画素値の比較画素の数が、予め設定された下位グループとする画素の個数より多ければ、処理対象画素S0が下位グループに属さないと判定する。そして、ノイズ判定部131は、処理対象画素S0が上位グループ及び下位グループのいずれにも属さない場合、処理対象画素S0が中位グループに属すると判定する。 Then, the noise determination unit 131 performs processing based on the number of comparison pixels having a pixel value larger than the processing target pixel S0 and the number of comparison pixels having a pixel value smaller than the processing target pixel S0, which is a count result in the comparison pixel group. It is determined whether each of the target pixel S0 and the comparison pixels S1 to S8 belongs to an upper group, a middle group, or a lower group (step S202). At this time, the noise determination unit 131 sets the pixels having the pixel values of the first and second ranks from the top as the upper group, and sets the pixels having the pixel values of the ninth and eighth ranks from the lower as the lower group. The pixels having the pixel values from the third position to the seventh position are set as a middle group. That is, the noise determination unit 131 compares the preset number of pixels as the upper group with the number of comparison pixels having a pixel value larger than the processing target pixel S0, and compares the pixels having a pixel value larger than the processing target pixel S0. Is smaller than a preset number of pixels for the upper group, it is determined that the processing target pixel S0 belongs to the upper group. On the other hand, the noise determination unit 131 determines that the processing target pixel S0 does not belong to the upper group if the number of comparison pixels having a pixel value larger than the processing target pixel S0 is larger than a preset number of pixels to be the upper group. To do. In addition, the noise determination unit 131 compares the number of pixels set as a lower group set in advance with the number of comparison pixels having a pixel value smaller than the processing target pixel S0, and the comparison pixel having a pixel value smaller than the processing target pixel S0. Is smaller than the number of pixels to be set as a lower group in advance, it is determined that the processing target pixel S0 belongs to the lower group. On the other hand, the noise determination unit 131 determines that the processing target pixel S0 does not belong to the lower group if the number of comparison pixels having a pixel value smaller than the processing target pixel S0 is larger than the number of pixels set as a lower group set in advance. To do. Then, when the processing target pixel S0 does not belong to either the upper group or the lower group, the noise determination unit 131 determines that the processing target pixel S0 belongs to the middle group.
 次に、ノイズ判定部131は、処理対象画素S0が上位グループまたは下位グループのいずれかに属しているか否かの判定を行う(ステップS203)。
 このとき、ノイズ判定部131は、処理対象画素S0が上位グループまたは下位グループのいずれかに属している場合、ノイズが重畳していると判定し、処理をステップS204へ進める。一方、ノイズ判定部131は、処理対象画素S0が中位グループに属している場合、ノイズが重畳していないと判定し、処理をステップS207へ進める。
Next, the noise determination unit 131 determines whether or not the processing target pixel S0 belongs to either the upper group or the lower group (step S203).
At this time, if the processing target pixel S0 belongs to either the upper group or the lower group, the noise determination unit 131 determines that noise is superimposed, and the process proceeds to step S204. On the other hand, when the processing target pixel S0 belongs to the middle group, the noise determination unit 131 determines that noise is not superimposed and advances the process to step S207.
 次に、ノイズ判定部131は、中位グループの画素値の平均値を求め(ステップS204)、処理対象画素S0の画素値と平均値との画素差分値を求め(ステップS205)、この画素差分値をノイズ値検出回路132へ出力する。 Next, the noise determination unit 131 obtains an average value of the pixel values of the middle group (step S204), obtains a pixel difference value between the pixel value and the average value of the processing target pixel S0 (step S205), and this pixel difference The value is output to the noise value detection circuit 132.
 次に、ノイズ判定部131は、処理対象画素S0が上位グループまたは下位グループのいずれかに属している場合、シフトレジスタSF2のレジスタSR25から出力される処理対象画素S0の画素値に、ノイズが処理対象画素S0に重畳していることを示すノイズフラグを付加し(ステップS206)、ノイズリダクション演算部133に対して出力する。 Next, when the processing target pixel S0 belongs to either the upper group or the lower group, the noise determination unit 131 processes noise in the pixel value of the processing target pixel S0 output from the register SR25 of the shift register SF2. A noise flag indicating that the pixel is superimposed on the target pixel S0 is added (step S206) and output to the noise reduction calculation unit 133.
 また、ノイズ判定部131は、処理対象画素S0が中位グループに属している場合において、ノイズが処理対象画素S0に重畳されていないことを示すノイズフラグを、シフトレジスタSF2のレジスタSR25から出力される画素値に付加して出力する(ステップS207)。 In addition, when the processing target pixel S0 belongs to the middle group, the noise determination unit 131 outputs a noise flag indicating that noise is not superimposed on the processing target pixel S0 from the register SR25 of the shift register SF2. Is added to the pixel value to be output (step S207).
 次に、ノイズリダクション演算部133は、ノイズフラグの値に対応し、供給される画素の画素値に対するノイズ除去の処理を行う(ステップS208)。
 すなわち、ノイズリダクション演算部133は、画素値に付加されたノイズフラグがノイズが処理対象画素S0に重畳されていないことを示す場合、供給された画素値をそのまま出力映像信号Soutとして出力する。
 一方、ノイズリダクション演算部133は、画素値に付加されたノイズフラグがノイズが重畳していることを示す場合、ノイズ値検出回路132から直前のフレームにおいて求めたノイズ量を読み込み、処理対象画素S0の画素値の補正を行う。このとき、ノイズリダクション演算部133は、ノイズフラグに上位グループを示すデータが付加されていると、処理対象画素S0の画素値からノイズ量を減算し、一方、ノイズフラグに下位グループを示すデータが付加されていると、処理対象画素S0の画素値に対してノイズ量を加算し、演算した結果を出力映像Soutとして出力する。
Next, the noise reduction calculation unit 133 performs noise removal processing on the pixel value of the supplied pixel corresponding to the value of the noise flag (step S208).
That is, when the noise flag added to the pixel value indicates that the noise is not superimposed on the processing target pixel S0, the noise reduction calculation unit 133 outputs the supplied pixel value as it is as the output video signal Sout.
On the other hand, when the noise flag added to the pixel value indicates that the noise is superimposed, the noise reduction calculation unit 133 reads the noise amount obtained in the previous frame from the noise value detection circuit 132, and processes the pixel S0 to be processed. The pixel value is corrected. At this time, if the data indicating the upper group is added to the noise flag, the noise reduction calculation unit 133 subtracts the noise amount from the pixel value of the processing target pixel S0, while the data indicating the lower group is included in the noise flag. If added, the amount of noise is added to the pixel value of the processing target pixel S0, and the calculated result is output as the output video Sout.
 そして、画素の画素値が入力される毎に、ステップS201からステップS208が繰り返され、各画素に対するノイズ除去処理が行われる。
 また、ノイズ値検出回路132は、ノイズ判定部131から入力される画素差分値の絶対値を求め、この絶対値を加算して加算絶対値を求め、垂直同期信号がY/C分離回路12から入力されると、加算絶対値を供給された画素差分値の数で除算し、画素差分値の絶対値の平均値を求め、ノイズ量とする。
 したがって、ノイズリダクション演算部133は、直前のフレームで求められたノイズ量を、現在のフレームの各画素のノイズ除去に用いている。ノイズリダクション演算部133は、垂直同期信号が入力される毎に、直前のフレームで得られた画素差分値から求めた新たなノイズ量をノイズ量レジスタに格納し、さらにその前のフレームで求めたノイズ量に上書きする。
Each time a pixel value of a pixel is input, steps S201 to S208 are repeated, and noise removal processing is performed on each pixel.
Further, the noise value detection circuit 132 obtains the absolute value of the pixel difference value input from the noise determination unit 131, adds the absolute value to obtain the addition absolute value, and the vertical synchronization signal is obtained from the Y / C separation circuit 12. When input, the added absolute value is divided by the number of supplied pixel difference values, and the average value of the absolute values of the pixel difference values is obtained to obtain the noise amount.
Therefore, the noise reduction calculation unit 133 uses the noise amount obtained in the immediately previous frame for noise removal of each pixel in the current frame. Each time the vertical synchronization signal is input, the noise reduction calculation unit 133 stores a new noise amount obtained from the pixel difference value obtained in the immediately preceding frame in the noise amount register, and further obtains it in the previous frame. Overwrite the amount of noise.
 また、上述した実施形態においては、処理対象画素S0が上位グループあるいは下位グループに属し、ノイズが重畳していると判定された場合、中位グループの画素の画素値の平均値を求め、処理対象画素S0の画素値からこの平均値を減算した画素差分値を、ノイズ量を求める際に用いている。
 しかしながら、ノイズ量を求める方法として、ノイズ判定部131は、処理対象画素S0が上位グループに含まれ、かつ処理対象画素S0の画素値が当該処理対象画素S0の画素値及び比較画素群に含まれる全ての比較画素の画素値のうちで最大値である場合と、処理対象画素S0が下位グループに含まれ、かつ処理対象画素S0の画素値が当該処理対象画素S0の画素値と比較画素群に含まれる全ての比較画素の画素値のうちで最小値である場合とにおいて、処理対象画素S0と他の画素との画素値の差分の平均値を求め、この平均値を画素差分値として、ノイズ値検出回路132へ出力するように構成しても良い。ノイズ値検出回路132のノイズ量を求める処理は上述した実施形態と同様である。
In the above-described embodiment, when it is determined that the processing target pixel S0 belongs to the upper group or the lower group and noise is superimposed, the average value of the pixel values of the pixels in the middle group is obtained, and the processing target A pixel difference value obtained by subtracting the average value from the pixel value of the pixel S0 is used when obtaining the noise amount.
However, as a method for obtaining the noise amount, the noise determination unit 131 includes the processing target pixel S0 in the upper group, and the pixel value of the processing target pixel S0 is included in the pixel value and the comparison pixel group of the processing target pixel S0. When the pixel value is the maximum value among the pixel values of all the comparison pixels, the processing target pixel S0 is included in the lower group, and the pixel value of the processing target pixel S0 is included in the comparison pixel group and the pixel value of the processing target pixel S0. In the case where the pixel value is the minimum value among the pixel values of all the comparison pixels included, an average value of pixel value differences between the processing target pixel S0 and the other pixels is obtained, and this average value is used as the pixel difference value. You may comprise so that it may output to the value detection circuit 132. FIG. The process of obtaining the noise amount of the noise value detection circuit 132 is the same as that in the above-described embodiment.
 また、さらに他のノイズ量を求める方法として、ノイズ判定部131は、比較画素群の画素とは異なる周辺の他の位置の複数の画素を予め第2の比較画素群として設定しておき、処理対象画素S0が上位グループに含まれた場合、処理対象画素S0の画素値がこの第2の比較画素群と比較して最大値または最小値である場合、処理対象画素S0と第2の比較画素群の画素の各々との画素値の差分の平均値を求め、この平均値を画素差分値として、ノイズ値検出回路132へ出力するように構成しても良い。ノイズ値検出回路132のノイズ量を求める処理は上述した実施形態と同様である。この第2の比較画素群の画素としては、例えば、シフトレジスタSF1におけるレジスタSR12、SR15からSR16、シフトレジスタSF2におけるレジスタSR22、SR25からSR26、シフトレジスタSF3におけるレジスタSR32、SR35からSR36の画素の組み合わせを用いる。 As another method for obtaining the amount of noise, the noise determination unit 131 sets a plurality of pixels at other positions in the vicinity different from the pixels of the comparison pixel group as a second comparison pixel group in advance, and performs processing. When the target pixel S0 is included in the upper group, when the pixel value of the processing target pixel S0 is the maximum value or the minimum value compared to the second comparison pixel group, the processing target pixel S0 and the second comparison pixel An average value of pixel value differences with respect to each of the pixels in the group may be obtained, and this average value may be output to the noise value detection circuit 132 as a pixel difference value. The process of obtaining the noise amount of the noise value detection circuit 132 is the same as that in the above-described embodiment. As the pixels of the second comparison pixel group, for example, the combination of the registers SR12 and SR15 to SR16 in the shift register SF1, the registers SR22 and SR25 to SR26 in the shift register SF2, and the registers SR32 and SR35 to SR36 in the shift register SF3. Is used.
 上述したように、本実施形態によれば、処理対象画素と、当該処理対象画素の周辺の画素とで比較画素群を構成し、処理対象画素がこの比較対象群のなかで、画素値が上位グループまたは下位グループのいずれかに属する場合にのみノイズ除去を実行するため、輪郭線に対応する処理対象画素にノイズが重畳されていると判断される確率を低くし、輪郭線をノイズとして不必要に消去して画像の惚けを生じさせることを低減させることで、画像の画質の低下を抑制したノイズ除去を行うことを可能とする。
 すなわち、本実施形態によれば、上述した画素値による画素のグルーピング処理により、上下左右斜めの如何なる方向に対する画像の輪郭線をも検出することができるため、輪郭線を不必要に消すことなく、画像のノイズ除去処理を行うことが可能となる。
As described above, according to the present embodiment, the processing target pixel and the pixels around the processing target pixel constitute a comparison pixel group, and the processing target pixel is in the comparison target group. Since noise removal is performed only when it belongs to either a group or a lower group, the probability that noise is superimposed on the processing target pixel corresponding to the contour line is reduced, and the contour line is unnecessary as noise. It is possible to perform noise removal while suppressing deterioration in image quality by reducing the occurrence of image blurring by erasing.
That is, according to the present embodiment, the pixel grouping process based on the pixel values described above can detect the contour line of the image in any direction up, down, left, and right, so that the contour line is not unnecessarily erased. Image noise removal processing can be performed.
 後述するように、本実施形態による画像処理装置を用いたノイズ除去の結果、特に画像における輪郭線の不必要な消去を低減させ、ぼけを抑制したノイズ除去が行えるという効果を奏することを確認した。
 図8は、図22の原画像に対して本実施形態によるノイズ除去を行った結果の画像を示す図である。図8の画像と図23の画像とを比較すると、図8の画像おけるノイズが図23の画像に対して低減されており、従来例に比較して本実施形態のノイズ除去の処理が、より効果的に行われていることが判る。
 また、図8の画像と図24の画像とを比較すると、図8の画像におけるノイズが図24の画像と同等程度に低減され、かつ図24の画像に比較してぼけが少なく(輪郭線が消去される程度が低く)、従来例に比較して本実施形態のノイズ除去の処理が、より効果的に行われていることが分かる。
 また、図9は、図25の原画像に対して本実施形態によるノイズ除去を行った結果の画像を示す図である。図9の画像と図26の画像とを比較すると、図26の画像に対して図9の画像の惚けが抑制され、図9の符号R1を付した領域内で色の違いにより認識できる窪みが図26に比較してはっきりしていることが判る。本実施形態及び後述する各実施形態の各々において、各実施形態の効果を説明する図8、図9、図13、図14、図18、図22、図23、図24、図25、図26は全てカラー画像であり、色の変化の度合いにより、惚けが抑制されていることが確認できる。
As will be described later, as a result of noise removal using the image processing apparatus according to the present embodiment, it has been confirmed that there is an effect that noise removal that suppresses blurring can be performed especially by reducing unnecessary erasing of an outline in an image. .
FIG. 8 is a diagram illustrating an image obtained as a result of noise removal performed on the original image of FIG. 22 according to the present embodiment. Comparing the image of FIG. 8 with the image of FIG. 23, the noise in the image of FIG. 8 is reduced with respect to the image of FIG. 23, and the noise removal processing of this embodiment is more effective than the conventional example. It turns out that it is done effectively.
Further, when the image of FIG. 8 is compared with the image of FIG. 24, the noise in the image of FIG. 8 is reduced to the same level as that of the image of FIG. 24, and there is less blur compared to the image of FIG. It can be seen that the noise removal processing of the present embodiment is performed more effectively than the conventional example.
FIG. 9 is a diagram illustrating an image obtained as a result of performing noise removal on the original image in FIG. 25 according to the present embodiment. When the image of FIG. 9 is compared with the image of FIG. 26, the blur of the image of FIG. 9 is suppressed with respect to the image of FIG. 26, and there is a dent that can be recognized by the difference in color in the region denoted by reference numeral R1 of FIG. As can be seen from FIG. 8, 9, 13, 14, 18, 18, 22, 23, 24, 25, and 26 for explaining the effect of each embodiment in this embodiment and each of the embodiments described later. Are all color images, and it can be confirmed that blurring is suppressed by the degree of color change.
<第2の実施形態>
 次に、本発明の第2の実施形態の説明を行う。第2の実施形態による液晶表示装置は、第1の実施形態における図1に示す構成と同様である。以下、第1の実施形態と同様の構成については同一の符号を付し、第1の実施形態と異なる構成及び動作(ノイズリダクション部の構成及び動作)のみについて説明する。
<Second Embodiment>
Next, a second embodiment of the present invention will be described. The liquid crystal display device according to the second embodiment has the same configuration as that shown in FIG. 1 in the first embodiment. Hereinafter, the same reference numerals are given to the same configurations as those in the first embodiment, and only configurations and operations different from those in the first embodiment (configurations and operations of the noise reduction unit) will be described.
 図10は、本実施形態におけるノイズ除去の概要を説明するための図である。同図において、Hは水平方向、Vは垂直方法、Fはフレーム方向である。また、フレーム間検出を、前後フレームの同一画素でピークを持つか調べ、前後フレームの周辺画素より大きいか調べることと定義する。
 本実施形態におけるノイズ除去において、ノイズリダクション部13がノイズ除去を行う際、処理対象の画素、すなわち処理対象画素にノイズが重畳しているか否かの判定を行う処理について説明する。
 ノイズリダクション部13(図1)には、液晶パネル16の液晶素子に横方向、すなわち主走査方向(ゲート線18方向)に順番に画素単位で、Y/C分離部12から供給される。そして、1行目の主走査線に配置された画素全てに対する映像信号が入力されると、2行目の主走査線に配置された画素の映像信号が入力され、順次の主走査線に対応する映像信号が入力される。また、1フレーム分の映像信号が供給されると、次のフレームの映像信号の供給が行われる。
FIG. 10 is a diagram for explaining an outline of noise removal in the present embodiment. In the figure, H is the horizontal direction, V is the vertical method, and F is the frame direction. Further, inter-frame detection is defined as checking whether the same pixel in the preceding and following frames has a peak and checking whether it is larger than the surrounding pixels in the preceding and following frames.
In the noise removal in this embodiment, when the noise reduction unit 13 performs noise removal, a process for determining whether noise is superimposed on a pixel to be processed, that is, a pixel to be processed will be described.
The noise reduction unit 13 (FIG. 1) is supplied from the Y / C separation unit 12 to the liquid crystal elements of the liquid crystal panel 16 in the horizontal direction, that is, in the main scanning direction (gate line 18 direction) in units of pixels. When the video signals for all the pixels arranged on the main scanning line in the first row are input, the video signals for the pixels arranged in the main scanning line on the second row are input, corresponding to the sequential main scanning lines. The video signal to be input is input. When a video signal for one frame is supplied, the video signal for the next frame is supplied.
 本実施形態においては、第1の実施形態と異なり、異なる複数のフレーム間において、各フレームの2次元平面(2次元空間)に配置された画素間の画素値を比較し(時間方向の比較)、画素にノイズが重畳しているか否かの判定を行う。
輝度信号Y、色差信号Cb、色差信号Crの各々に対して独立に、ノイズ除去の処理を行うが、第1の実施形態と同様に、説明を簡単にするため、これらを画素値として以下説明する。第2の実施形態においては、各フレームの2次元平面と、この2次元平面の前後のフレームとの組み合わせを、画素が配列された画素空間としてノイズ除去の処理を行う。
 例えば、図10において示すように、処理対象画素S0が位置するフレームF2と、処理対象画素S0の位置するフレームF0に対して直前のフレームF1及び直後のフレームF2とに位置する画素の組み合わせにより比較画素群を構成する。
In the present embodiment, unlike the first embodiment, pixel values between pixels arranged in a two-dimensional plane (two-dimensional space) of each frame are compared between different frames (time direction comparison). Then, it is determined whether noise is superimposed on the pixel.
Noise removal processing is performed independently for each of the luminance signal Y, the color difference signal Cb, and the color difference signal Cr. However, in the same manner as in the first embodiment, these will be described below as pixel values in order to simplify the description. To do. In the second embodiment, noise removal processing is performed using a combination of a two-dimensional plane of each frame and frames before and after the two-dimensional plane as a pixel space in which pixels are arranged.
For example, as shown in FIG. 10, the frame F2 where the processing target pixel S0 is located and the frame F0 where the processing target pixel S0 is located are compared by a combination of pixels located in the immediately preceding frame F1 and the immediately following frame F2. A pixel group is configured.
 ここで、比較画素群を構成する比較画素として、フレームF1において、フレームF0における処理対象画素S0に対応する画素の周辺の画素S10、S11、S12、S13、S14、S15、S16、S17及びS18を設定している。
 また、比較画素群を構成する比較画素として、フレームF2において、フレームF0における処理対象画素S0に対応する画素の周辺の画素S20、S21、S22、S23、S24、S25、S26、S27及びS28を設定している。
 ここで、フレームF0における画素S0に対し、フレームF1の画素S10とフレームF2の画素S20とは、表示画面のフレームにおいては同一の位置である。同様に、フレームF1における画素S11からS18の各々と、フレームF2における画素S21からS28の各々とも、表示画面のフレームにおいては同一の位置である。
Here, as comparison pixels constituting the comparison pixel group, in the frame F1, pixels S10, S11, S12, S13, S14, S15, S16, S17, and S18 around the pixel corresponding to the processing target pixel S0 in the frame F0 are set. It is set.
In addition, in the frame F2, pixels S20, S21, S22, S23, S24, S25, S26, S27, and S28 around the pixel corresponding to the processing target pixel S0 in the frame F0 are set as the comparison pixels constituting the comparison pixel group. is doing.
Here, with respect to the pixel S0 in the frame F0, the pixel S10 in the frame F1 and the pixel S20 in the frame F2 are in the same position in the frame of the display screen. Similarly, each of the pixels S11 to S18 in the frame F1 and each of the pixels S21 to S28 in the frame F2 are in the same position in the frame of the display screen.
 したがって、本実施形態において、ノイズリダクション部13は、フレームF0における比較対象画素S0と、フレームF1における比較画素S10からS18と、フレームF2における比較画素S20からS28との、17個の画素により比較画素群を構成している。ノイズリダクション部13は、この比較画素群における17個の画素において、比較対象画素S0と、比較画素との画素値の大小を判定し、第1の実施形態と同様に、比較対象画素S0より画素値が大きい比較画素の個数と、比較対象画素S0より画素値が小さい比較画素の個数とを各々計数する。 Therefore, in the present embodiment, the noise reduction unit 13 compares the comparison pixel S0 with the 17 pixels of the comparison target pixel S0 in the frame F0, the comparison pixels S10 to S18 in the frame F1, and the comparison pixels S20 to S28 in the frame F2. It constitutes a group. The noise reduction unit 13 determines the magnitude of the pixel value of the comparison target pixel S0 and the comparison pixel among the 17 pixels in the comparison pixel group, and the pixel from the comparison target pixel S0 is the same as in the first embodiment. The number of comparison pixels having a large value and the number of comparison pixels having a pixel value smaller than that of the comparison target pixel S0 are counted.
 そして、第1の実施形態と同様に、ノイズリダクション部13は、比較対象画素S0より画素値が大きい比較画素の個数と、比較対象画素S0より画素値が小さい比較画素の個数とに基づき、各画素を上位グループ、中位グループ及び下位グループの3つのグループに振り分ける。
 ここで、ノイズリダクション部13は、処理対象画素S0が上位グループまたは下位グループに属した場合、この処理対象画素S0にノイズが重畳されていると判定し、中位グループに属した場合、この処理対象画素S0にノイズが重畳されていないと判定する。
Then, as in the first embodiment, the noise reduction unit 13 determines each pixel based on the number of comparison pixels having a pixel value larger than the comparison target pixel S0 and the number of comparison pixels having a pixel value smaller than the comparison target pixel S0. Pixels are divided into three groups: an upper group, a middle group, and a lower group.
Here, when the processing target pixel S0 belongs to the upper group or the lower group, the noise reduction unit 13 determines that noise is superimposed on the processing target pixel S0. When the processing target pixel S0 belongs to the middle group, this process is performed. It is determined that no noise is superimposed on the target pixel S0.
 本実施形態においては、上述した順位の配列の上位5個の画素(最上位から5個目までの1位から5位の画素)を上位グループとし、順位の配列の下位5個の画素(最下位から5個目までの15位から19位の画素)を下位グループとし、残りの8画素(6位から14位の画素)を中位グループとする。
 すでに説明したように、画像における図形の輪郭線は、図4に示すように、垂直方向(x軸方向)の輪郭線L1、水平方向(y軸方向)の輪郭線L2、斜め方向の輪郭線L3及び曲線形状の輪郭線L4などのように、最低3個の、すなわち処理対象画素及び比較画素の合計が3個以上の画素から構成される場合が多いと考えられる。すなわち、処理対象画素が1個であるため、3個以上の比較画素を設定し、比較画素群を4個とする必要がある。
In the present embodiment, the top five pixels (first to fifth pixels from the top to the fifth) in the above-described rank array are set as a top group, and the bottom five pixels (highest) in the rank array. The pixels from the 15th to the 19th pixel from the lowest to the fifth) are the lower group, and the remaining 8 pixels (the 6th to 14th pixels) are the middle group.
As already described, as shown in FIG. 4, the contour lines of the figure in the image are the contour line L1 in the vertical direction (x-axis direction), the contour line L2 in the horizontal direction (y-axis direction), and the contour line in the oblique direction. Like L3 and the contour line L4 having a curved shape, it is considered that there are many cases where at least three pixels, that is, the total of the processing target pixel and the comparison pixel is composed of three or more pixels. That is, since there is one processing target pixel, it is necessary to set three or more comparison pixels and to set four comparison pixel groups.
 したがって、19個の画素からなる比較画素群において、フレーム間における画素の位置関係は、比較画素及び処理対象画素の位置がそれぞれ、各フレームにおいて同一の位置であるから、輪郭線が直前のフレームで3個、処理対象画素の位置するフレームで1個、直後のフレームで3個の計7個から検出されることになる。これにより、処理対象画素S0が輪郭線に含まれている場合、すなわち輪郭線が存在していれば、7個以上の画素で構成されることになり、処理対象画素S0が6位以下で14位以内の中位グループに属する確率が高くなるため、ノイス除去の処理が行われる確率も低くなる。 Therefore, in the comparison pixel group consisting of 19 pixels, the positional relationship of the pixels between frames is that the positions of the comparison pixel and the processing target pixel are the same in each frame, so that the contour line is the previous frame. The detection is made from a total of seven, three, one in the frame where the pixel to be processed is located and three in the immediately following frame. Thereby, when the processing target pixel S0 is included in the contour line, that is, when the contour line exists, the processing target pixel S0 is composed of seven or more pixels. Since the probability of belonging to the middle group within the rank is increased, the probability that the noise removal process is performed is also decreased.
 上述したように、処理対象画素S0が輪郭線の画素である場合、中位グループに属する可能性が高くなる。したがって、ノイズリダクション部13は、上位グループあるいは下位グループに属する場合、画素にノイズが重畳していると判定する。
 また、さらに処理対象画素S0に対するノイズ除去の処理が行われる確率を低くする場合、上位グループ及び下位グループをそれぞれ3個とし、中位グループを13個の画素で構成する。このことにより、上述した上位グループ及び下位グループが5個の場合に比較して、処理対象画素S0が中位グループに属する確率をより高くすることもできる。
As described above, when the processing target pixel S0 is a contour pixel, the possibility of belonging to the middle group increases. Therefore, the noise reduction unit 13 determines that noise is superimposed on the pixel when belonging to the upper group or the lower group.
Further, in order to further reduce the probability that noise removal processing is performed on the processing target pixel S0, the upper group and the lower group are each three, and the middle group is configured by 13 pixels. Accordingly, it is possible to further increase the probability that the processing target pixel S0 belongs to the middle group as compared with the case where the above-described upper group and lower group are five.
 また、上述の説明においては、比較画素群を、処理対象画素を含めて19画素で説明したが、この画素数及びその画素数とした場合の上位グループ及び下位グループとする画素の数は、第1の実施形態と同様に、どのような環境で、どのような画像を送信するかにより、予め複数の組み合わせにて実験を行い、ノイズ除去の理結果から設定する。
 また、上位グループと下位グループとの画素の個数は同一の数にする必要がなく、上述したように、グループの画素数の決定の際の組み合わせから、ノイズ除去の処理結果に応じて設定する。ただし、処理対象画素S0の位置するフレームF0と、直前のフレームF1と直後のフレームF2において、輪郭線を検出する必要性から、各フレームにおける比較画素の位置は同一である、すなわち比較する比較画素がフレーム間で同一である。
In the above description, the comparison pixel group has been described as 19 pixels including the processing target pixel. However, when the number of pixels and the number of pixels are used, the number of pixels as the upper group and the lower group is as follows. As in the first embodiment, an experiment is performed in advance in a plurality of combinations depending on what kind of image and what kind of image is transmitted, and the result of noise removal is set.
Further, the number of pixels in the upper group and the lower group does not need to be the same, and as described above, the number of pixels in the group is set according to the processing result of noise removal from the combination when determining the number of pixels. However, in the frame F0 where the processing target pixel S0 is located, and in the immediately preceding frame F1 and the immediately following frame F2, the position of the comparison pixel in each frame is the same because of the necessity of detecting the contour line, that is, the comparison pixel to be compared. Are the same between frames.
 また、比較画素群における処理対象画素と比較する比較画素の位置も、上述した予め行う処理結果により設定する。このため、比較画素の位置は、処理対象画素に対して対称の位置にある必要はない。
 例えば、図10のフレームF1において、フレームF0の比較画素群S0に対応する画素S10の周辺の他の画素の位置は、画素S10と画素S15とは3画素離れ、処理対象画素S10と画素S11とは1画素離れ、画素S10と画素S17は4画素離れ、画素S10と画素S13とは2画素離れている。フレームF2においても同様に、処理対象画素S0に対応する画素S20に対し、画素空間において比較画素として選択される周辺の画素は、画素S20に対して対象ではない。
Further, the position of the comparison pixel to be compared with the processing target pixel in the comparison pixel group is also set by the above-described processing result performed in advance. For this reason, the position of the comparison pixel does not have to be symmetrical with respect to the processing target pixel.
For example, in the frame F1 of FIG. 10, the positions of the other pixels around the pixel S10 corresponding to the comparison pixel group S0 of the frame F0 are 3 pixels apart from the pixel S10 and the pixel S11. Is one pixel apart, pixel S10 and pixel S17 are four pixels apart, and pixel S10 and pixel S13 are two pixels apart. Similarly, in the frame F2, the peripheral pixels selected as the comparison pixels in the pixel space with respect to the pixel S20 corresponding to the processing target pixel S0 are not targets for the pixel S20.
 次に、図11を用いて図1におけるノイズリダクション部13の説明を行う。図11は図1におけるノイズリダクション部13の構成例を示すブロック図である。この図11は、図10に示す以下に示す18画素と処理対象画素S0とにより構成した比較画素群に対応した回路構成となっている。ここで、フレームF1において、フレームF0の比較画素群S0に対応する画素S10の周辺の他の画素の位置は、画素S10と画素S15とは3画素離れ、処理対象画素S10と画素S11とは1画素離れ、画素S10と画素S17は4画素離れ、画素S10と画素S13とは2画素離れている。また、フレームF2において、フレームF0の比較画素群S0に対応する画素S20の周辺の他の画素の位置は、画素S20と画素S25とは3画素離れ、処理対象画素S20と画素S21とは1画素離れ、画素S20と画素S27は4画素離れ、画素S20と画素S23とは2画素離れている。
 第1の実施形態と同様に、このノイズリダクション部13は、輝度信号Yと、色差信号Cbと、色差信号Crとの各々に対して設けられている。以下、これら輝度信号Yと、色差信号Cbと、色差信号Crの信号の数値を、画素毎の画素値として説明する。
Next, the noise reduction unit 13 in FIG. 1 will be described with reference to FIG. FIG. 11 is a block diagram showing a configuration example of the noise reduction unit 13 in FIG. FIG. 11 shows a circuit configuration corresponding to a comparison pixel group constituted by the following 18 pixels and the processing target pixel S0 shown in FIG. Here, in the frame F1, the positions of other pixels around the pixel S10 corresponding to the comparison pixel group S0 of the frame F0 are separated by 3 pixels from the pixel S10 and the pixel S15, and the processing target pixel S10 and the pixel S11 are 1 Pixel separation, pixel S10 and pixel S17 are four pixels apart, and pixel S10 and pixel S13 are two pixels apart. In the frame F2, the positions of other pixels around the pixel S20 corresponding to the comparison pixel group S0 of the frame F0 are separated by 3 pixels from the pixel S20 and the pixel S25, and the processing target pixel S20 and the pixel S21 are 1 pixel. The pixel S20 and the pixel S27 are separated by 4 pixels, and the pixel S20 and the pixel S23 are separated by 2 pixels.
Similar to the first embodiment, the noise reduction unit 13 is provided for each of the luminance signal Y, the color difference signal Cb, and the color difference signal Cr. Hereinafter, numerical values of the luminance signal Y, the color difference signal Cb, and the color difference signal Cr will be described as pixel values for each pixel.
 ノイズリダクション部13は、フレームメモリFM1、FM2、ラインメモリLM11、LM12、LM21、LM31、LM32、シフトレジスタSF11、SF12、SF13、SF21、SF31、SF32、SF33、ノイズ判定部131、ノイズ値検出部132、ノイズリダクション演算部133及び制御部134とを備えている。例えば、各フレームメモリ、ラインメモリ、シフトレジスタは、映像が256階調のため、Y、Cr、Cbの各々の構成において、8ビットとなる。 The noise reduction unit 13 includes frame memories FM1, FM2, line memories LM11, LM12, LM21, LM31, LM32, shift registers SF11, SF12, SF13, SF21, SF31, SF32, SF33, a noise determination unit 131, and a noise value detection unit 132. , A noise reduction calculation unit 133 and a control unit 134 are provided. For example, each frame memory, line memory, and shift register has 8 bits in each of Y, Cr, and Cb because the video has 256 gradations.
 制御部134は、Y/C分離部12から供給される垂直同期信号により、新たなフレームの画素値の処理が開始されるか否かの判定を行うとともに、フレームFM1、FM2、ラインメモリLM11、LM12、LM21、LM31及びLM32と、シフトレジスタSF11、SF12、SF13、SF21、SF31、SF32及びSF33の各々における画素値の入力処理及びシフト処理を行う。 The control unit 134 determines whether or not the processing of the pixel value of the new frame is started based on the vertical synchronization signal supplied from the Y / C separation unit 12, and the frames FM1 and FM2, the line memory LM11, Pixel value input processing and shift processing are performed in each of the LM12, LM21, LM31, and LM32, and the shift registers SF11, SF12, SF13, SF21, SF31, SF32, and SF33.
 フレームメモリFM1及びFM1は、1フレームを構成する全画素の画素値を記憶する容量を有している。
 ラインメモリLM11及びLM31は、3主走査線に配置された液晶素子PIXの数の画素値を直列に記憶する容量を有している。一方、ラインメモリLM12、LM21及びLM32は、5走査線分の容量を有している。
 図11において、入力される映像信号Sinにおける画素は、2つ前のフレームの画素がフレームメモリFM1に記憶され、直前のフレームの画素がフレームFM1に記憶されている。このフレームメモリFM1及びFM2の各々は、1フレームの画素の画素値が順次転送される直列のシフトレジスタとして形成されている。このため、映像信号Sinとして入力される画素の画素値は、順次フレームメモリFM2及びFM1からなる、2フレーム分の画素の画素値を転送するシフトレジスタにおいて、順次制御部134により転送される。そして、ラインメモリLM21には、フレームメモリFM2からフレームメモリFM1へ転送される画素値が順次入力される。また、ラインメモリLM12には、フレームメモリFM1から出力される画素値が順次入力される。
 また、本実施形態において、ノイズ除去を行う処理対象画素S0は、シフトレジスタSF21におけるレジスタSF215の画素値の画素である。
The frame memories FM1 and FM1 have a capacity for storing pixel values of all the pixels constituting one frame.
The line memories LM11 and LM31 have a capacity for storing in series the pixel values of the number of liquid crystal elements PIX arranged on the three main scanning lines. On the other hand, the line memories LM12, LM21, and LM32 have a capacity for five scanning lines.
In FIG. 11, as for the pixels in the input video signal Sin, the pixel of the previous frame is stored in the frame memory FM1, and the pixel of the immediately previous frame is stored in the frame FM1. Each of the frame memories FM1 and FM2 is formed as a serial shift register to which pixel values of one frame of pixels are sequentially transferred. For this reason, the pixel values of the pixels input as the video signal Sin are sequentially transferred by the control unit 134 in the shift register that sequentially transfers the pixel values of the pixels for two frames including the frame memories FM2 and FM1. Then, the pixel values transferred from the frame memory FM2 to the frame memory FM1 are sequentially input to the line memory LM21. Further, the pixel values output from the frame memory FM1 are sequentially input to the line memory LM12.
Further, in the present embodiment, the processing target pixel S0 from which noise is removed is a pixel having a pixel value of the register SF215 in the shift register SF21.
 すなわち、フレームメモリFM1からnフレーム目が出力され、フレームメモリFM2からn+1フレーム目が出力され、映像信号Sinとしてn+2フレーム目が入力される時点からノイズ除去の処理が開始される。
 ここで、nフレームであるフレームF1の比較画素と、n+2フレームであるF2の比較画素と、n+1フレームであるF0の処理対象画素S0とにより構成される比較画素対照群にて、シフトレジスタSF21のレジスタSF215の画素値(処理対象画素S0の画素値)のノイズ除去の処理が行われる。
That is, the nth frame is output from the frame memory FM1, the (n + 1) th frame is output from the frame memory FM2, and the noise removal process is started when the n + 2th frame is input as the video signal Sin.
Here, in the comparison pixel contrast group constituted by the comparison pixel of the frame F1 which is n frame, the comparison pixel of F2 which is n + 2 frame, and the processing target pixel S0 of F0 which is n + 1 frame, the shift register SF21 Noise removal processing is performed on the pixel value of the register SF215 (pixel value of the processing target pixel S0).
 また、シフトレジスタSF11、SF12、SF13、SF21、SF31、SF32及びSF33の各々は、8ビット幅のレジスタが7段直列に接続されたシフトレジスタであり、図10に示す比較画素群における処理対象画素S0に対する比較画素S10からS18、S20からS28の各々の配置位置を設定するものである。
 ここで、シフトレジスタSF11におけるレジスタSF111、SF115、S117と、シフトレジスタSF12におけるレジスタSF121、SF125、S127と、シフトレジスタSF13におけるレジスタSF131、SF135、S137とからは、それぞれフレームF1の画素S16、S17、S18、S15、S10、S11、S14、S13、S12の各々の画素値が、ノイズ判定部131へ入力される。
 また、シフトレジスタSF21におけるレジスタSF215からは、処理対象画素S0の画素値が、ノイズ判定部131へ入力される。
 また、シフトレジスタSF31におけるレジスタSF311、SF315、S317と、シフトレジスタSF32におけるレジスタSF321、SF325、S327と、シフトレジスタSF33におけるレジスタSF331、SF335、S337とからは、それぞれフレームF2の画素S26、S27、S28、S25、S20、S21、S24、S23、S22の各々の画素値が、ノイズ判定部131へ入力される。
Further, each of the shift registers SF11, SF12, SF13, SF21, SF31, SF32, and SF33 is a shift register in which 8-bit width registers are connected in series in seven stages, and the pixel to be processed in the comparison pixel group shown in FIG. The arrangement positions of the comparison pixels S10 to S18 and S20 to S28 for S0 are set.
Here, the registers SF111, SF115, and S117 in the shift register SF11, the registers SF121, SF125, and S127 in the shift register SF12, and the registers SF131, SF135, and S137 in the shift register SF13, respectively, are the pixels S16, S17, and S17 in the frame F1, respectively. The pixel values of S18, S15, S10, S11, S14, S13, and S12 are input to the noise determination unit 131.
Further, the pixel value of the processing target pixel S0 is input to the noise determination unit 131 from the register SF215 in the shift register SF21.
The registers SF311, SF315, and S317 in the shift register SF31, the registers SF321, SF325, and S327 in the shift register SF32, and the registers SF331, SF335, and S337 in the shift register SF33 are the pixels S26, S27, and S28 of the frame F2, respectively. , S25, S20, S21, S24, S23, and S22 are input to the noise determination unit 131.
 ノイズ判定部131は、処理対象画素S0と、比較画素S10からS18及びS20からS28の19の画素の各々との画素値の大小判定を行い、処理対象画素S0より画素値の大きい比較画素の個数と、処理対象画素S0より画素値の小さい比較画素の個数とを、各々計数する。
 また、ノイズ判定部131は、配列において順位の高い5個(最大値の順位1位から順位5位)の5画素を上位グループとし、配列において順位の低い5個(最低値の順位15位と順位19位)の5画素を下位グループとし、配列における6位から14位の9個の画素を中位グループとする。
The noise determination unit 131 determines the pixel value of the processing target pixel S0 and each of the 19 pixels of the comparison pixels S10 to S18 and S20 to S28, and the number of comparison pixels having a pixel value larger than that of the processing target pixel S0. And the number of comparison pixels having a pixel value smaller than that of the processing target pixel S0 are counted.
In addition, the noise determination unit 131 sets the five pixels having the highest rank in the array (the highest rank from the first rank to the fifth rank) as the upper group, and the five pixels having the lower rank in the array (the lowest rank of the 15th rank). 5 pixels in the order 19) are set as a lower group, and 9 pixels from the 6th position to the 14th position in the array are set as a middle group.
 さらに、ノイズ判定部131は、処理対象画素S0より画素値が大きい比較画素の個数が4個以内の場合、処理対象画素S0を含む5画素を上位グループとし、一方、処理対象画素S0より画素値が小さい比較画素の個数が4個以内の場合を、処理対象画素S0を含む5画素を下位グループとする。そして、ノイズ判定部131は、処理対象画素S0が上位グループあるいは下位グループのいずれかに属している場合、この処理対象画素S0にノイズが重畳されていると判定する。そして、ノイズ判定部131は、上位グループ及び下位グループのいずれにも属さない9個の画素から構成される中位グループにおける画素値の平均値を算出し、処理対象画素S0の画素値からこの平均値を減算し、減算して得られた画素差分値をノイズ値検出回路132へ出力する。
 また、ノイズ判定部131は、処理対象画素S0にノイズが重畳されている場合、この処理対象画素S0にノイズが重畳していることを示すノイズフラグ(上位グループまたは下位グループのいずれかに属するかを示す情報も含む)を付加し、ノイズリダクション演算部133へ出力する。
 ここで、ノイズ判定部131は、ノイズリダクション演算部133に対して、シフトレジスタSF21におけるレジスタSF213の画素値に、ノイズフラグを付加して順次出力する。
Further, when the number of comparison pixels having a pixel value larger than that of the processing target pixel S0 is four or less, the noise determination unit 131 sets the five pixels including the processing target pixel S0 as an upper group, while the pixel value of the processing target pixel S0 is higher than that of the processing target pixel S0. When the number of comparison pixels having a small value is 4 or less, 5 pixels including the processing target pixel S0 are set as a lower group. Then, when the processing target pixel S0 belongs to either the upper group or the lower group, the noise determination unit 131 determines that noise is superimposed on the processing target pixel S0. Then, the noise determination unit 131 calculates an average value of pixel values in the middle group composed of nine pixels that do not belong to either the upper group or the lower group, and calculates this average from the pixel value of the processing target pixel S0. The pixel difference value obtained by subtracting the value is output to the noise value detection circuit 132.
In addition, when noise is superimposed on the processing target pixel S0, the noise determination unit 131 indicates a noise flag indicating whether noise is superimposed on the processing target pixel S0 (whether it belongs to either the upper group or the lower group) Is added to the noise reduction calculation unit 133.
Here, the noise determination unit 131 adds the noise flag to the pixel value of the register SF213 in the shift register SF21 and sequentially outputs it to the noise reduction calculation unit 133.
 ノイズ値検出回路132は、ノイズ判定部131から供給される、ノイズが重畳していると判定された処理対象画素S0に対応した画素差分値の絶対値を求め、この画素差分値の絶対値を1フレーム分加算し、供給された画素差分値の数により加算結果を除算することにより、ノイズ値を算出する。 The noise value detection circuit 132 obtains the absolute value of the pixel difference value corresponding to the processing target pixel S0 supplied from the noise determination unit 131 and determined to have noise superimposed thereon, and calculates the absolute value of this pixel difference value. The noise value is calculated by adding one frame and dividing the addition result by the number of supplied pixel difference values.
 ノイズリダクション演算部133は、シフトレジスタSF25から順次供給される処理対象画素S0の画素値に、ノイズが重畳していることを示すノイズフラグが付加されている場合、ノイズ値検出回路132が全フレームにおいて算出したノイズ値を読み込む。
 そして、ノイズリダクション演算部133は、処理対象画素S0が上位グループに属していることをノイズフラグが示している場合、供給される画素値からノイズ値を減算する演算を行い、また、処理対象画素S0が下位グループに属していることをノイズフラグが示している場合、供給される画素値にノイズ値を加算する演算を行い、演算結果をノイズ除去処理を行った映像信号Soutとして出力する。
 一方、ノイズリダクション演算部133は、重畳されていないことを示すノイズフラグが画素値に付加されている場合、ノイズ判定部131から供給される画素値をそのまま、映像信号Soutとして出力する。
When a noise flag indicating that noise is superimposed is added to the pixel value of the processing target pixel S0 sequentially supplied from the shift register SF25, the noise reduction calculation unit 133 causes the noise value detection circuit 132 to perform the entire frame. The noise value calculated in is read.
When the noise flag indicates that the processing target pixel S0 belongs to the upper group, the noise reduction calculation unit 133 performs a calculation to subtract the noise value from the supplied pixel value, and the processing target pixel When the noise flag indicates that S0 belongs to the lower group, an operation of adding the noise value to the supplied pixel value is performed, and the operation result is output as a video signal Sout subjected to noise removal processing.
On the other hand, when a noise flag indicating that they are not superimposed is added to the pixel value, the noise reduction calculation unit 133 outputs the pixel value supplied from the noise determination unit 131 as it is as the video signal Sout.
 次に、図12のフローチャートは、第2の実施形態による画像処理装置のノイズリダクション部13におけるノイズ除去の処理の動作例を示す。また、画像処理装置の表示装置1全体としての処理としては、図6に示すフローチャートと同様のため、第1の実施形態と異なるノイズリダクション部13の動作のみを以下に説明する。
 また、図12に示す動作は、Y/C分離部12から供給される垂直同期信号により、制御部134が新たなフレームの画素値の処理が開始されるか否かの判定を行い開始される。
すなわち、制御部134は、フレームメモリFM1、FM2、ラインメモリLM11、LM12、LM21、LM31及びLM32と、シフトレジスタSF11、SF12、SF13、SF21、SF31、SF32、SF33とに対し、映像信号Sinとしての画素値の入力処理を行う。
Next, the flowchart of FIG. 12 shows an operation example of the noise removal processing in the noise reduction unit 13 of the image processing apparatus according to the second embodiment. Since the processing of the entire display device 1 of the image processing apparatus is the same as the flowchart shown in FIG. 6, only the operation of the noise reduction unit 13 different from the first embodiment will be described below.
The operation shown in FIG. 12 is started by determining whether or not the processing of the pixel value of the new frame is started by the control unit 134 based on the vertical synchronization signal supplied from the Y / C separation unit 12. .
That is, the control unit 134 outputs the video signals Sin to the frame memories FM1 and FM2, the line memories LM11, LM12, LM21, LM31 and LM32, and the shift registers SF11, SF12, SF13, SF21, SF31, SF32 and SF33. Performs pixel value input processing.
 制御部134は、電源の入力があるか、またはチャネルが変更され、そして同一チャネルにおいてノイズ除去が開始されてから、入力されるフレームがmフレーム以降、本実施形態においては例えば3フレーム以降であるか否かの判定を行う(ステップS300)。
このとき、制御部134は、同一チャネルにおいてノイズ除去の処理が開始されてから、Y/C分離部12から供給される垂直同期信号の数をカウントし、垂直同期信号カウント数として内部の記憶部に書き込んで記憶し、この垂直同期カウント数が3以上であるか否かの判定を行っている、すなわち、映像信号が比較画素群を構成するために必要な3枚のフレームが入力されているか否かの判定を行う。また、制御部134は、この垂直同期カウント数を、電源が入力あるいはチャネルが変更される際にリセットし、「0」とする。
 そして、制御部134は、垂直同期カウント数が3以上である場合、処理をステップS301へ進め、一方、水平同期カウント数が3未満である場合、処理をステップS307へ進める。
The control unit 134 receives m frames after the power is input or the channel is changed and noise removal is started in the same channel, and in this embodiment, for example, 3 frames or more in this embodiment. Is determined (step S300).
At this time, the control unit 134 counts the number of vertical synchronization signals supplied from the Y / C separation unit 12 after the start of noise removal processing in the same channel, and stores the internal synchronization unit as the vertical synchronization signal count number. Is written and stored, and it is determined whether or not the vertical synchronization count number is 3 or more, that is, whether the three frames necessary for the video signal to form the comparison pixel group are input. Determine whether or not. Further, the control unit 134 resets the vertical synchronization count number to “0” when the power is input or the channel is changed.
Then, when the vertical synchronization count number is 3 or more, the control unit 134 advances the process to step S301, and when the horizontal synchronization count number is less than 3, the control unit 134 advances the process to step S307.
 本実施形態においては、フレームF1及びF2における処理対象画素に対応する画素S10、S20のある主走査線に隣接する上下の主走査線にある画素を比較画素としたが、処理対象画素S0に対応する画素S10、S20のある主走査線から複数本の主走査線を介した主走査線を、処理対象画素のある主走査線に対して上下で設定しても良い。 In this embodiment, the pixels on the upper and lower main scanning lines adjacent to the main scanning line having the pixels S10 and S20 corresponding to the processing target pixels in the frames F1 and F2 are used as the comparison pixels. The main scanning lines from the main scanning line having the pixels S10 and S20 through the plurality of main scanning lines may be set up and down with respect to the main scanning line having the processing target pixel.
 また、フレームF1及びF2において、処理対象画素と比較する画素が、処理対象画素に対応する位置にある画素S10、S20に対して対象の位置にある必要はなく、かつ、処理対象画素のある主走査線に対し、処理対象画素に対応する位置にある画素S10、S20のある主走査線に対して対象の位置にある必要はない。比較画素群を構成するために必要な、処理対象画素の存在する主走査線に対し、処理対象画素と比較する画素の主走査線との位置関係及び各主走査線における画素の位置関係を、用いる対象の画像の種類(変化の激しい映像、あまり変化のない映像)などにより、それぞれ実験により設定する。 Further, in the frames F1 and F2, the pixel to be compared with the processing target pixel does not need to be at the target position with respect to the pixels S10 and S20 at the position corresponding to the processing target pixel, and the main pixel having the processing target pixel is present. The scanning line need not be at the target position with respect to the main scanning line having the pixels S10 and S20 at the position corresponding to the processing target pixel. The positional relationship between the processing target pixel and the main scanning line of the pixel to be processed and the positional relationship of the pixel in each main scanning line with respect to the main scanning line where the processing target pixel exists, which is necessary to configure the comparison pixel group, Depending on the type of target image to be used (video that changes rapidly, video that does not change much), etc., each is set by experiment.
 次に、制御部134は、ノイズ判定部131に対して、処理対象画素S0にノイズが重畳しているか否かの判定を行うことを指示する制御信号を出力する。
 そして、ノイズ判定部131は、シフトレジスタSF11のレジスタSF111、SF115及びSF117と、シフトレジスタSF12のレジスタSF121、SF125及びSF127と、シフトレジスタSF13のレジスタSF131、SF135及びSF137と、シフトレジスタSF21のレジスタSF215と、シフトレジスタSF31のレジスタSF311、SF315及びSF317と、シフトレジスタSF32のレジスタSF321、SF325及びSF327と、シフトレジスタSF33のレジスタSF331、SF335及びSF337との各々から、画素S26、S27、S28、S25、S20、S21、S24、S23、S22、処理対象画素S0、画素S16、S17、S18、S15、S10、S11、S14、S13、S12の画素値を入力する。
Next, the control unit 134 outputs a control signal that instructs the noise determination unit 131 to determine whether noise is superimposed on the processing target pixel S0.
The noise determination unit 131 includes the registers SF111, SF115, and SF117 of the shift register SF11, the registers SF121, SF125, and SF127 of the shift register SF12, the registers SF131, SF135, and SF137 of the shift register SF13, and the register SF215 of the shift register SF21. Each of the registers SF311, SF315, and SF317 of the shift register SF31, the registers SF321, SF325, and SF327 of the shift register SF32, and the registers SF331, SF335, and SF337 of the shift register SF33, and the pixels S26, S27, S28, S25, S20, S21, S24, S23, S22, processing target pixel S0, pixels S16, S17, S18, S15, S10, S11, S 4, S13, and inputs the pixel values of the S12.
 次に、ノイズ判定部131は、処理対象画素S0と、画素S10から画素S18、画素S20から画素S28の各々とで画素値を比較し、処理対象画素S0より画素値の大きい比較画素の個数と、処理対象画素S0より画素値の小さい比較画素の個数との各々の計数を行う(ステップS301)。 Next, the noise determination unit 131 compares the pixel values of the processing target pixel S0 with each of the pixels S10 to S18 and the pixels S20 to S28, and determines the number of comparison pixels having a pixel value larger than that of the processing target pixel S0. The number of comparison pixels having a pixel value smaller than that of the processing target pixel S0 is counted (step S301).
 そして、ノイズ判定部131は、配列した処理対象画素S0、画素S10から画素S18、画素20から画素28の各々の画素値が、上位グループ、中位グループまたは下位グループのいずれのグループに属するかの判定を、第1実施形態と同様に行う(ステップS302)。ここで、ノイズ判定部131は、上からの順位が1位から5位の画素値の5個の画素を上位グループとし、下位から順位が19位から15位の画素値の5個の画素を下位グループとし、残りの6位から14位の画素値の9個の画素を中位グループとする。
 このとき、ノイズ判定部131は、予め設定された上位グループとする画素の個数と、処理対象画素S0より大きい画素値の比較画素の個数の5個とを比較し、処理対象画素S0より大きい画素値の比較画素の数が、予め設定された上位グループとする画素の個数である5個より少なければ、処理対象画素S0が上位グループに属すると判定する。一方、ノイズ判定部131は、処理対象画素S0より大きい画素値の比較画素の数が、予め設定された上位グループとする画素の個数より多ければ、処理対象画素S0が上位グループに属さないと判定する。また、ノイズ判定部131は、予め設定された下位グループとする画素の個数と、処理対象画素S0より小さい画素値の比較画素の個数とを比較し、処理対象画素S0より小さい画素値の比較画素の数が、予め設定された下位グループとする画素の個数が5個より少なければ、処理対象画素S0が下位グループに属すると判定する。一方、ノイズ判定部131は、処理対象画素S0より小さい画素値の比較画素の数が、予め設定された下位グループとする画素の個数の5個より多ければ、処理対象画素S0が下位グループに属さないと判定する。そして、ノイズ判定部131は、処理対象画素S0が上位グループ及び下位グループのいずれにも属さない場合、処理対象画素S0が中位グループに属すると判定する。
Then, the noise determination unit 131 determines whether the pixel values of the arranged processing target pixels S0, pixels S10 to S18, and pixels 20 to 28 belong to any of the upper group, middle group, or lower group. The determination is performed in the same manner as in the first embodiment (step S302). Here, the noise determination unit 131 sets the five pixels having the pixel values of the first to fifth ranks from the top as the upper group and the five pixels having the pixel values of the 19th to 15th ranks from the lower. The lower group is set, and the remaining nine pixels having the pixel values of the sixth to the 14th are set as the middle group.
At this time, the noise determination unit 131 compares the preset number of pixels in the upper group with five of the number of comparison pixels having a pixel value larger than the processing target pixel S0, and the pixels larger than the processing target pixel S0. If the number of comparison pixels of the value is less than 5 which is a preset number of pixels for the upper group, it is determined that the processing target pixel S0 belongs to the upper group. On the other hand, the noise determination unit 131 determines that the processing target pixel S0 does not belong to the upper group if the number of comparison pixels having a pixel value larger than the processing target pixel S0 is larger than a preset number of pixels to be the upper group. To do. In addition, the noise determination unit 131 compares the number of pixels set as a lower group set in advance with the number of comparison pixels having a pixel value smaller than the processing target pixel S0, and the comparison pixel having a pixel value smaller than the processing target pixel S0. If the number of pixels is less than 5 as a preset lower group, it is determined that the processing target pixel S0 belongs to the lower group. On the other hand, if the number of comparison pixels having a pixel value smaller than the processing target pixel S0 is greater than the preset number of pixels of the lower group, the noise determination unit 131 belongs to the lower group. Judge that there is no. Then, when the processing target pixel S0 does not belong to either the upper group or the lower group, the noise determination unit 131 determines that the processing target pixel S0 belongs to the middle group.
 次に、ノイズ判定部131は、処理対象画素S0が上位グループまたは下位グループのいずれかに属しているか否かの判定を行う(ステップS303)。
 このとき、ノイズ判定部131は、処理対象画素S0が上位グループまたは下位グループのいずれかに属している場合、ノイズが重畳していると判定し、処理をステップS304へ進める。一方、ノイズ判定部131は、処理対象画素S0が中位グループに属している場合、ノイズが重畳していないと判定し、処理をステップS307へ進める。
Next, the noise determination unit 131 determines whether or not the processing target pixel S0 belongs to either the upper group or the lower group (step S303).
At this time, if the processing target pixel S0 belongs to either the upper group or the lower group, the noise determination unit 131 determines that noise is superimposed, and the process proceeds to step S304. On the other hand, if the processing target pixel S0 belongs to the middle group, the noise determination unit 131 determines that noise is not superimposed, and advances the process to step S307.
 次に、ノイズ判定部131は、中位グループの画素の画素値の平均値を求め(ステップS304)、処理対象画素S0の画素値と平均値との画素差分値を求め(ステップS305)、この画素差分値をノイズ値検出回路132へ出力する。 Next, the noise determination unit 131 obtains an average value of the pixel values of the middle group pixels (step S304), obtains a pixel difference value between the pixel value and the average value of the processing target pixel S0 (step S305), and The pixel difference value is output to the noise value detection circuit 132.
 次に、ノイズ判定部131は、シフトレジスタSF21のレジスタSF213から出力される処理対象画素S0の画素値に、ノイズが重畳していることを示すノイズフラグを付加し(ステップS306)、ノイズリダクション演算部133に対して出力する。 Next, the noise determination unit 131 adds a noise flag indicating that noise is superimposed on the pixel value of the processing target pixel S0 output from the register SF213 of the shift register SF21 (step S306), and performs noise reduction calculation. Output to the unit 133.
 また、ノイズ判定部131は、ステップS300において処理対象画素S0に対する比較画素群が構成できないと判定された(ノイズが重畳しているか否かの判定ができない)場合と、処理対象画素S0が中位グループに属している場合とにおいて、ノイズが処理対象画素に重畳されていないことを示すノイズフラグを、シフトレジスタSF21のレジスタSF215から出力される画素値に付加して出力する(ステップS307)。 The noise determination unit 131 determines that the comparison pixel group for the processing target pixel S0 cannot be configured in step S300 (cannot determine whether noise is superimposed), and the processing target pixel S0 is in the middle level. In the case of belonging to a group, a noise flag indicating that noise is not superimposed on the processing target pixel is added to the pixel value output from the register SF215 of the shift register SF21 and output (step S307).
 次に、ノイズリダクション演算部133は、ノイズフラグの値に対応し、供給される画素の画素値に対するノイズ除去の処理を行う(ステップS308)。
 すなわち、ノイズリダクション演算部133は、画素値に付加されたノイズフラグがノイズが処理対象画素に重畳されていないことを示す場合、供給された画素値をそのまま出力映像信号Soutとして出力する。
 一方、ノイズリダクション演算部133は、画素値に付加されたノイズフラグがノイズが重畳していることを示す場合、ノイズ値検出回路132から直前のフレームにおいて求めたノイズ量を読み込み、処理対象画素S0の画素値の補正を行う。このとき、ノイズリダクション演算部133は、ノイズフラグに上位グループを示すデータが付加されていると、処理対象画素S0の画素値からノイズ量を減算し、一方、ノイズフラグに下位グループを示すデータが付加されていると、処理対象画素S0の画素値に対してノイズ量を加算し、演算した結果を出力映像Soutとして出力する。
Next, the noise reduction calculation unit 133 performs noise removal processing on the pixel value of the supplied pixel corresponding to the value of the noise flag (step S308).
That is, when the noise flag added to the pixel value indicates that the noise is not superimposed on the processing target pixel, the noise reduction calculation unit 133 outputs the supplied pixel value as it is as the output video signal Sout.
On the other hand, when the noise flag added to the pixel value indicates that the noise is superimposed, the noise reduction calculation unit 133 reads the noise amount obtained in the previous frame from the noise value detection circuit 132, and processes the pixel S0 to be processed. The pixel value is corrected. At this time, if the data indicating the upper group is added to the noise flag, the noise reduction calculation unit 133 subtracts the noise amount from the pixel value of the processing target pixel S0, while the data indicating the lower group is included in the noise flag. If added, the amount of noise is added to the pixel value of the processing target pixel S0, and the calculated result is output as the output video Sout.
 そして、画素の画素値が入力される毎に、ステップS301からステップS308が繰り返され、各画素に対するノイズ除去処理が行われる。
 また、ノイズ値検出回路132は、ノイズ判定部131から入力される画素差分値の絶対値を求め、この絶対値を加算して加算絶対値を求め、垂直同期信号がY/C分離回路12から入力されると、加算絶対値を供給された画素差分値の数で除算し、画素差分値の絶対値の平均値を求め、ノイズ量とする。
 したがって、ノイズリダクション演算部133は、第1の実施形態と同様に、直前のフレームで求められたノイズ量を、現在のフレームの各画素のノイズ除去に用いている。ノイズリダクション演算部133は、垂直同期信号が入力される毎に、直前のフレームで得られた画素差分値から求めた新たなノイズ量をノイズ量レジスタに格納し、さらにその前のフレームで求めたノイズ量に上書きする。
Each time a pixel value of a pixel is input, steps S301 to S308 are repeated, and noise removal processing is performed on each pixel.
Further, the noise value detection circuit 132 obtains the absolute value of the pixel difference value input from the noise determination unit 131, adds the absolute value to obtain the addition absolute value, and the vertical synchronization signal is obtained from the Y / C separation circuit 12. When input, the added absolute value is divided by the number of supplied pixel difference values, and the average value of the absolute values of the pixel difference values is obtained to obtain the noise amount.
Therefore, as in the first embodiment, the noise reduction calculation unit 133 uses the noise amount obtained in the immediately preceding frame for noise removal of each pixel in the current frame. Each time the vertical synchronization signal is input, the noise reduction calculation unit 133 stores a new noise amount obtained from the pixel difference value obtained in the immediately preceding frame in the noise amount register, and further obtains it in the previous frame. Overwrite the amount of noise.
 また、上述した実施形態においては、処理対象画素S0が上位グループあるいは下位グループに属し、ノイズが重畳していると判定された場合、中位グループの画素の画素値の平均値を求め、処理対象画素S0の画素値からこの平均値を減算した画素差分値を、ノイズ量を求める際に用いている。
 しかしながら、ノイズ量を求める方法として、ノイズ判定部131は、処理対象画素S0が上位グループに含まれ、かつ画素値が比較画素群において最大値である場合と、処理対象画素S0が下位グループに含まれ、かつ画素値が最低値である場合とにおいて、処理対象画素S0と他の画素との画素値の差分の平均値を求め、この平均値を画素差分値として、ノイズ値検出回路132へ出力するように構成しても良い。ノイズ値検出回路132のノイズ量を求める処理は上述した実施形態と同様である。
In the above-described embodiment, when it is determined that the processing target pixel S0 belongs to the upper group or the lower group and noise is superimposed, the average value of the pixel values of the pixels in the middle group is obtained, and the processing target A pixel difference value obtained by subtracting the average value from the pixel value of the pixel S0 is used when obtaining the noise amount.
However, as a method of obtaining the noise amount, the noise determination unit 131 includes the case where the processing target pixel S0 is included in the upper group and the pixel value is the maximum value in the comparison pixel group, and the processing target pixel S0 is included in the lower group. In the case where the pixel value is the lowest value, the average value of the difference between the pixel values of the processing target pixel S0 and the other pixels is obtained, and this average value is output to the noise value detection circuit 132 as the pixel difference value. You may comprise so that it may do. The process of obtaining the noise amount of the noise value detection circuit 132 is the same as that in the above-described embodiment.
 また、さらに他のノイズ量を求める方法として、ノイズ判定部131は、比較画素群の画素とは異なる周辺の他の位置の複数の画素を予め第2の比較画素群として設定しておき、処理対象画素S0が上位グループに含まれた場合、処理対象画素S0の画素値がこの第2の比較画素群と比較して最大値である場合と、処理対象画素S0が下位グループに含まれ、かつ画素値が最低値である場合とにおいて、処理対象画素S0と第2の比較画素群の画素の各々との画素値の差分の平均値を求め、この平均値を画素差分値として、ノイズ値検出回路132へ出力するように構成しても良い。ノイズ値検出回路132のノイズ量を求める処理は上述した実施形態と同様である。 As another method for obtaining the amount of noise, the noise determination unit 131 sets a plurality of pixels at other positions in the vicinity different from the pixels of the comparison pixel group as a second comparison pixel group in advance, and performs processing. When the target pixel S0 is included in the upper group, the pixel value of the processing target pixel S0 is the maximum value compared to the second comparison pixel group, the processing target pixel S0 is included in the lower group, and When the pixel value is the lowest value, an average value of pixel value differences between the processing target pixel S0 and each pixel of the second comparison pixel group is obtained, and this average value is used as the pixel difference value to detect the noise value. It may be configured to output to the circuit 132. The process of obtaining the noise amount of the noise value detection circuit 132 is the same as that in the above-described embodiment.
 例えば、第2の比較画素群の画素としては、シフトレジスタSF11のレジスタSF112、SF115及びSF116と、シフトレジスタSF12のレジスタSF122、SF125及びSF126と、シフトレジスタSF13のレジスタSF132、SF135及びSF136と、シフトレジスタSF31のレジスタSF312、SF315及びSF316と、シフトレジスタSF32のレジスタSF322、SF325及びSF326と、シフトレジスタSF33のレジスタSF332、SF335及びSF336との画素の組み合わせを用いる。 For example, the pixels of the second comparison pixel group include the registers SF112, SF115, and SF116 of the shift register SF11, the registers SF122, SF125, and SF126 of the shift register SF12, the registers SF132, SF135, and SF136 of the shift register SF13, and the shift A combination of pixels of the registers SF312, SF315, and SF316 of the register SF31, the registers SF322, SF325, and SF326 of the shift register SF32, and the registers SF332, SF335, and SF336 of the shift register SF33 is used.
 上述したように、本実施形態によれば、処理対象画素の属するフレームF0の前後のフレームF1及びF2において、フレームF0における処理対象画素S0と同一の位置にある画素と、この画素の周辺の画素と、処理対象画素S0とで比較画素群を構成するため、動きのある映像であるとしても、フレームF0にある処理対象画素S0の画素と同一の画素値を有する画素がフレームF1及びF2のいずれか、あるいは双方において同一の位置に存在する確率は高いと考えられる。 As described above, according to the present embodiment, in the frames F1 and F2 before and after the frame F0 to which the processing target pixel belongs, the pixel at the same position as the processing target pixel S0 in the frame F0 and the pixels around this pixel Since the comparison pixel group is composed of the processing target pixel S0, even if the image has a motion, a pixel having the same pixel value as the pixel of the processing target pixel S0 in the frame F0 is displayed in any of the frames F1 and F2. Or it is considered that the probability of being in the same position in both is high.
 したがって、フレームF1、F2及びF3の3つのフレームにおいて、処理対象画素S0を含み、輪郭線を構成する画素が7個であることを含め、それぞれ5個の画素から構成される上位グループまたは下位グループに属する場合、ノイズが重畳している確率が高く、一方中位グループに属する場合、ノイズが重畳していない確率が高いと考えられる。そして、処理対象画素S0がこの比較対象群のなかで、画素値が上位グループまたは下位グループのいずれかに属する場合にのみノイズ除去を実行するため、輪郭線に対応する処理対象画素S0にノイズが重畳されていると判断される確率を低くし、輪郭線をノイズとして不必要に消去して画像の惚けを低減させることで、画像の画質の低下を抑制したノイズ除去を行うことを可能とする。
 すなわち、本実施形態によれば、上述した画素値による画素のグルーピング処理により、上下左右斜めの如何なる方向に対する画像の輪郭線をも検出することができるため、輪郭線を不必要に消すことなく、画像のノイズ除去処理を行うことが可能となる。
Accordingly, in the three frames F1, F2 and F3, the upper group or the lower group each including five pixels including the processing target pixel S0 and including seven pixels constituting the contour line. If the signal belongs to the middle group, the probability that the noise is not superimposed is high. Since the noise removal is executed only when the processing target pixel S0 is in the comparison target group and the pixel value belongs to either the upper group or the lower group, the processing target pixel S0 corresponding to the contour line has noise. It is possible to perform noise removal that suppresses degradation of image quality by reducing the probability of being judged as being superposed and unnecessarily erasing the outline as noise to reduce image blurring. .
That is, according to the present embodiment, the pixel grouping process based on the pixel values described above can detect the contour line of the image in any direction up, down, left, and right, so that the contour line is not unnecessarily erased. Image noise removal processing can be performed.
 図13は、図22の原画像に対して本実施形態によるノイズ除去を行った結果の画像を示す図である。図13の画像と図24の画像とを比較すると、図13の画像におけるノイズが図24の画像に比較して低減され、かつ図24の画像に比較してぼけが少なく(輪郭線が消去される程度が低く)、従来例に比較して本実施形態のノイズ除去の処理が、より効果的に行われていることが判る。
 また、図14は、図25の原画像に対して本実施形態によるノイズ除去を行った結果の画像を示す図である。図14の画像と図26の画像とを比較すると、図26画像に対して図14の画像のぼけが抑制され、図14の符号R1を付した領域内で色の違いにより認識できる窪みが図26の画像に比較してはっきりしていることが判る。
FIG. 13 is a diagram illustrating an image obtained as a result of noise removal performed on the original image in FIG. 22 according to the present embodiment. Comparing the image of FIG. 13 with the image of FIG. 24, the noise in the image of FIG. 13 is reduced compared to the image of FIG. 24, and there is less blur compared to the image of FIG. It can be seen that the noise removal processing of the present embodiment is performed more effectively than the conventional example.
FIG. 14 is a diagram illustrating an image obtained as a result of performing noise removal on the original image in FIG. 25 according to the present embodiment. Comparing the image of FIG. 14 with the image of FIG. 26, the blur of the image of FIG. 14 is suppressed with respect to the image of FIG. 26, and a dent that can be recognized by the difference in color in the region denoted by reference numeral R1 of FIG. It can be seen that it is clearer than the 26 images.
<第3の実施形態>
 次に、本発明の第3の実施形態の説明を行う。第3の実施形態による液晶表示装置は、第1の実施形態における図1に示す構成と同様である。以下、第1の実施形態と同様の構成については同一の符号を付し、第1の実施形態と異なる構成及び動作(ノイズリダクション部の構成及び動作)のみについて説明する。図15は、第3の実施形態によるノイズリダクション部13の構成例を示す図である。また、本実施形態による図1におけるノイズリダクション部13は、図15に示すように、第1ノイズリダクション部13Aと第2ノイズリダクション部13Bとから構成されている。すなわち、第1ノイズリダクション部13Aは、入力される映像信号Sinにおける画素の画素値に対して第1ノイズ除去処理を行い、このノイズ除去の結果を映像信号Soとして出力する。
 また、第2ノイズリダクション部13Bは、第1ノイズリダクション部13Aから出力される映像信号Soの画素値に対して第2ノイズ除去処理を行い、このノイズ除去の結果を映像信号Soutとして出力する。
<Third Embodiment>
Next, a third embodiment of the present invention will be described. The liquid crystal display device according to the third embodiment has the same configuration as that shown in FIG. 1 in the first embodiment. Hereinafter, the same reference numerals are given to the same configurations as those in the first embodiment, and only configurations and operations different from those in the first embodiment (configurations and operations of the noise reduction unit) will be described. FIG. 15 is a diagram illustrating a configuration example of the noise reduction unit 13 according to the third embodiment. Moreover, the noise reduction part 13 in FIG. 1 by this embodiment is comprised from 13 A of 1st noise reduction parts, and the 2nd noise reduction part 13B, as shown in FIG. That is, the first noise reduction unit 13A performs the first noise removal process on the pixel value of the pixel in the input video signal Sin, and outputs the result of the noise removal as the video signal So.
The second noise reduction unit 13B performs a second noise removal process on the pixel value of the video signal So output from the first noise reduction unit 13A, and outputs the result of the noise removal as a video signal Sout.
 次に、図16は、第1ノイズリダクション部13Aの構成例を示す図である。この第1ノイズリダクション部13Aは、第2の実施形態における図11のノイズリダクション部13と同様の構成については同一の符号を付してある。この第1ノイズリダクション部13Aと第2の実施形態におけるノイズリダクション部13と異なる点は、ノイズ量をヒストグラムから求めている点である。
 ノイズ判定部131Aは、第2の実施形態で説明した処理により得られた画素差分値を、ヒストグラム分析回路135Aに出力する。このノイズ判定部131Aは、他の動作については第2の実施形態におけるノイズ判定部131と同様である。
 また、制御回路134Aは、第2の実施形態における図11の制御部134と同様である。
Next, FIG. 16 is a diagram illustrating a configuration example of the first noise reduction unit 13A. In this first noise reduction unit 13A, the same components as those of the noise reduction unit 13 of FIG. 11 in the second embodiment are denoted by the same reference numerals. The difference between the first noise reduction unit 13A and the noise reduction unit 13 in the second embodiment is that the amount of noise is obtained from a histogram.
The noise determination unit 131A outputs the pixel difference value obtained by the processing described in the second embodiment to the histogram analysis circuit 135A. The noise determination unit 131A is the same as the noise determination unit 131 in the second embodiment in other operations.
The control circuit 134A is the same as the control unit 134 of FIG. 11 in the second embodiment.
 ヒストグラム分析部135Aは、供給される画素差分値の絶対値を求め、この画素差分値の絶対値が複数の数値範囲のいずれの範囲に属するかを判定し、範囲毎に属する画素差分値の数を積算して、1フレーム単位毎にヒストグラムを生成する。
 そして、ヒストグラム分析部135Aは、Y/C分離部12から垂直同期信号が入力される毎に、生成したヒストグラムにおける最も画素差分値数の多い数値範囲、すなわち画素差分値の最頻値を選択し、選択した最頻値を新たなノイズ量としてノイズ量レジスタに上書きして格納する。
The histogram analysis unit 135A obtains the absolute value of the supplied pixel difference value, determines which range of the plurality of numerical ranges the absolute value of the pixel difference value belongs, and the number of pixel difference values belonging to each range And a histogram is generated for each frame unit.
Then, each time a vertical synchronization signal is input from the Y / C separation unit 12, the histogram analysis unit 135A selects a numerical range having the largest number of pixel difference values in the generated histogram, that is, a mode value of pixel difference values. The selected mode value is overwritten and stored in the noise amount register as a new noise amount.
 ここで、上記数値範囲は、例えば、実験値により求めた画素差分値の絶対値の最大値から0までを複数の数値範囲に分割し、最も頻度の高い数値範囲の下側の数値をノイズ量とし、ノイズ量から実験値により求めた画素差分値の絶対値の最大値までを一つの数値範囲とする。
 具体的には、実験で得られた画素差分値の絶対値の最大値が120である場合でも、ノイズ量が30なら1から30まで、すなわち1、2、3、…、15、…、29、30とした1差違いのカウンタを設け、31以上はまとめて1つのカウンタとする。すなわち、画素差分値の数値範囲は、1から30までが30分割されて1であり、31以上は31となる。
 そして、ノイズリダクション演算部133Aは、第2の実施形態のノイズリダクション演算部133と同様に、直前のフレームで求めたノイズ量を用いて、処理対象画素S0のノイズ除去の処理に用いる。このノイズリダクション演算部133Aは、他の動作については第2の実施形態におけるノイズリダクション演算部133と同様である。
Here, the numerical range is divided into a plurality of numerical ranges, for example, from the maximum absolute value of pixel difference values obtained by experimental values to 0, and the lower numerical value in the most frequent numerical range is the noise amount. And a numerical value range from the noise amount to the maximum absolute value of the pixel difference value obtained from the experimental value.
Specifically, even if the maximum value of the absolute value of the pixel difference value obtained by the experiment is 120, if the noise amount is 30, 1 to 30, that is, 1, 2, 3,..., 15,. , 30 are provided as one-difference counters, and 31 or more are combined into one counter. That is, the numerical range of the pixel difference value is 1 by dividing 30 from 1 to 30, and 31 is 31 or more.
Then, the noise reduction calculation unit 133A is used for noise removal processing of the processing target pixel S0 using the noise amount obtained in the immediately preceding frame, similarly to the noise reduction calculation unit 133 of the second embodiment. The noise reduction calculation unit 133A is the same as the noise reduction calculation unit 133 in the second embodiment in other operations.
 次に、図17は、第2ノイズリダクション部13Bの構成例を示す図である。この第2ノイズリダクション部13Bは、第1の実施形態における図5のノイズリダクション部13と同様の構成については同一の符号を付してある。この第2ノイズリダクション部13Bと第1の実施形態におけるノイズリダクション部13と異なる点は、ノイズ量をヒストグラムから求めている点である。
 ノイズ判定部131Bは、第1の実施形態で説明した処理により得られた画素差分値を、ヒストグラム分析回路135Bに出力する。このノイズ判定部131Bは、他の動作については第1の実施形態におけるノイズ判定部131と同様である。
 また、制御回路134Bは、第1の実施形態における図5の制御部134と同様である。
Next, FIG. 17 is a diagram illustrating a configuration example of the second noise reduction unit 13B. In the second noise reduction unit 13B, the same components as those of the noise reduction unit 13 in FIG. 5 in the first embodiment are denoted by the same reference numerals. The difference between the second noise reduction unit 13B and the noise reduction unit 13 in the first embodiment is that the amount of noise is obtained from a histogram.
The noise determination unit 131B outputs the pixel difference value obtained by the processing described in the first embodiment to the histogram analysis circuit 135B. The noise determination unit 131B is the same as the noise determination unit 131 in the first embodiment in other operations.
The control circuit 134B is the same as the control unit 134 of FIG. 5 in the first embodiment.
 ヒストグラム分析部135Bは、供給される画素差分値の絶対値を求め、この画素差分値の絶対値が複数の数値範囲のいずれの範囲に属するかを判定し、範囲毎に属する画素差分値の数を積算して、1フレーム単位毎にヒストグラムを生成する。
 そして、ヒストグラム分析部135Bは、Y/C分離部12から垂直同期信号が入力される毎に、生成したヒストグラムにおける最も画素差分値数の多い数値範囲、すなわち画素差分値の最頻値を選択し、この数値範囲に対応するノイズ量を、新たなノイズ量をノイズ量レジスタに上書きして格納する。
 ここで、上記数値範囲は、すでに第1ノイズリダクション部13Aで説明したノイズ量の関係と同様である。
The histogram analysis unit 135B obtains the absolute value of the supplied pixel difference value, determines which range of the plurality of numerical ranges the absolute value of the pixel difference value belongs, and the number of pixel difference values belonging to each range And a histogram is generated for each frame unit.
Then, each time a vertical synchronization signal is input from the Y / C separation unit 12, the histogram analysis unit 135B selects a numerical range with the largest number of pixel difference values in the generated histogram, that is, a mode value of pixel difference values. The noise amount corresponding to this numerical range is stored by overwriting the new noise amount in the noise amount register.
Here, the numerical range is the same as the relationship of the noise amount already described in the first noise reduction unit 13A.
 上述したように、第3の実施形態は、ノイズ量の算出が異なるのみの第2の実施形態によるノイズリダクション部13と、第1の実施形態によるノイズリダクション部13とを直列に接続し、映像信号Sinの画素の画素値に対するノイズ除去を行う。
 このため、第3の実施形態も、第1及び第2の実施形態において説明したように、画層の輪郭線をぼやかすことなく、画像の質を低下させずにノイズ除去の処理を行うことができる。
 図18は、図22の原画像に対して本実施形態によるノイズ除去を行った結果の画像を示す図である。図18の画像と図24の画像とを比較すると、図18の画像におけるノイズが図24の画像に比較して低減され、かつ図24の画像に比較してぼけが少なく(輪郭線が消去される程度が低く)、従来例に比較して本実施形態のノイズ除去の処理が、より効果的に行われていることが判る。
As described above, in the third embodiment, the noise reduction unit 13 according to the second embodiment, which is different only in the calculation of the amount of noise, and the noise reduction unit 13 according to the first embodiment are connected in series, and video Noise removal is performed on the pixel value of the pixel of the signal Sin.
For this reason, in the third embodiment, as described in the first and second embodiments, the noise removal process is performed without blurring the outline of the layer and without degrading the image quality. Can do.
FIG. 18 is a diagram illustrating an image obtained as a result of performing noise removal on the original image in FIG. 22 according to the present embodiment. Comparing the image of FIG. 18 with the image of FIG. 24, the noise in the image of FIG. 18 is reduced compared to the image of FIG. 24, and there is less blur compared to the image of FIG. It can be seen that the noise removal processing of the present embodiment is performed more effectively than the conventional example.
 最後に、ノイズリダクション部13の各ブロックは、集積回路(ICチップ)上に形成された論理回路によって信号処理装置というハードウェアとして構成されてもよいし、ラインメモリ及びフレームメモリ以外の構成を次のようにCPU(Central Processing Unit)を用いてソフトウェアによって実現されてもよい。また、表示装置は、有機EL表示装置、CRT(陰極線管)表示装置、等であってもよい。 Finally, each block of the noise reduction unit 13 may be configured as hardware called a signal processing device by a logic circuit formed on an integrated circuit (IC chip), and the configuration other than the line memory and the frame memory is as follows. As described above, it may be realized by software using a CPU (Central Processing Unit). The display device may be an organic EL display device, a CRT (cathode ray tube) display device, or the like.
 すなわち、図5及び図11の各々におけるノイズ判定部131、ノイズ値検出部132、ノイズリダクション演算部133、制御部134と、図16におけるノイズ判定部131A、ノイズリダクション演算部133A、制御部134A、ヒストグラム分析部135Aと、図17におけるノイズ判定部131B、ノイズリダクション演算部133B、制御部134B、ヒストグラム分析部135Bとの機能を実現するためのプログラムをコンピュータ読み取り可能な記録媒体に記録して、この記録媒体に記録されたプログラムをコンピュータシステムに読み込ませ、実行することにより画像のノイズ除去の処理を行ってもよい。なお、ここでいう「コンピュータシステム」とは、OSや周辺機器等のハードウェアを含むものとする。 That is, the noise determination unit 131, the noise value detection unit 132, the noise reduction calculation unit 133, and the control unit 134 in FIG. 5 and FIG. 11, and the noise determination unit 131A, the noise reduction calculation unit 133A, the control unit 134A in FIG. A program for realizing the functions of the histogram analysis unit 135A, the noise determination unit 131B, the noise reduction calculation unit 133B, the control unit 134B, and the histogram analysis unit 135B in FIG. 17 is recorded on a computer-readable recording medium. Image noise removal processing may be performed by causing a computer system to read and execute a program recorded on a recording medium. Here, the “computer system” includes an OS and hardware such as peripheral devices.
 また、「コンピュータシステム」は、WWWシステムを利用している場合であれば、ホームページ提供環境(あるいは表示環境)も含むものとする。
 また、「コンピュータ読み取り可能な記録媒体」とは、フレキシブルディスク、光磁気ディスク、ROM、CD-ROM等の可搬媒体、コンピュータシステムに内蔵されるハードディスク等の記憶装置のことをいう。さらに「コンピュータ読み取り可能な記録媒体」とは、インターネット等のネットワークや電話回線等の通信回線を介してプログラムを送信する場合の通信線のように、短時間の間、動的にプログラムを保持するもの、その場合のサーバやクライアントとなるコンピュータシステム内部の揮発性メモリのように、一定時間プログラムを保持しているものも含むものとする。また上記プログラムは、前述した機能の一部を実現するためのものであっても良く、さらに前述した機能をコンピュータシステムにすでに記録されているプログラムとの組み合わせで実現できるものであっても良い。
Further, the “computer system” includes a homepage providing environment (or display environment) if a WWW system is used.
The “computer-readable recording medium” refers to a storage device such as a flexible medium, a magneto-optical disk, a portable medium such as a ROM or a CD-ROM, and a hard disk incorporated in a computer system. Furthermore, the “computer-readable recording medium” dynamically holds a program for a short time like a communication line when transmitting a program via a network such as the Internet or a communication line such as a telephone line. In this case, a volatile memory in a computer system serving as a server or a client in that case, and a program that holds a program for a certain period of time are also included. The program may be a program for realizing a part of the functions described above, and may be a program capable of realizing the functions described above in combination with a program already recorded in a computer system.
 以上、この発明の実施形態を図面を参照して詳述してきたが、具体的な構成はこの実施形態に限られるものではなく、この発明の要旨を逸脱しない範囲の設計等も含まれる。 As described above, the embodiment of the present invention has been described in detail with reference to the drawings. However, the specific configuration is not limited to this embodiment, and includes design and the like within the scope not departing from the gist of the present invention.
 11…検波部
 12…Y/C分離部
 13…ノイズリダクション部
 14…RGB変換部
 15…液晶駆動部
 16…液晶パネル
 18…ゲート線
 19…ソース線
131,131A,131B…ノイズ判定部
132…ノイズ値検出回路
133,133A,133B…ノイズリダクション演算部
134,134A,134B…制御部
FM1,FM2…フレームメモリ
LM1,LM2,LM11,LM12,LM13,LM21,LM31,LM32,LM33…ラインメモリ
SF1,SF2,SF3,SF11,SF12,SF13,SF21,SF31,SF32,SF33…シフトレジスタ
SR11,SR12,SR13,SR14,SR15,SR16,SR17,SR21,SR22,SR23,SR24,SR25,SR26,SR27,SR31,SR32,SR33,SR34,SR35,SR36,SR37…レジスタ
SF111,SF112,SF113,SF114,SF115,SF116,SF117,SF121,SF122,SF123,SF124,SF125,SF126,SF127,SF131,SF132,SF133,SF134,SF135,SF136,SF137,SF211,SF212,SF213,SF214,SF215,SF216,SF217,SF311,SF312,SF313,SF314,SF315,SF316,SF317,SF321,SF322,SF323,SF324,SF325,SF326,SF327,SF331,SF332,SF333,SF334,SF335,SF336,SF337…レジスタ
S0…処理対象画素
S1,S2,S3,S4,S5,S6,S7,S8,S10,S11,S12,S13,S14,S15,S16,S17,S18,S,20,S21,S22,S23,S24,S25,S26,S27,S28…比較画素
PIX…液晶素子
DESCRIPTION OF SYMBOLS 11 ... Detection part 12 ... Y / C separation part 13 ... Noise reduction part 14 ... RGB conversion part 15 ... Liquid crystal drive part 16 ... Liquid crystal panel 18 ... Gate line 19 ... Source line 131, 131A, 131B ... Noise determination part 132 ... Noise Value detection circuits 133, 133A, 133B ... noise reduction calculation units 134, 134A, 134B ... control units FM1, FM2 ... frame memories LM1, LM2, LM11, LM12, LM13, LM21, LM31, LM32, LM33 ... line memories SF1, SF2 , SF3, SF11, SF12, SF13, SF21, SF31, SF32, SF33 ... shift registers SR11, SR12, SR13, SR14, SR15, SR16, SR17, SR21, SR22, SR23, SR24, SR25, SR26, SR27, S R31, SR32, SR33, SR34, SR35, SR36, SR37 ... registers SF111, SF112, SF113, SF114, SF115, SF116, SF117, SF121, SF122, SF123, SF124, SF125, SF126, SF127, SF131, SF132, SF133, SF134 , SF135, SF136, SF137, SF211, SF212, SF213, SF214, SF215, SF216, SF217, SF311, SF312, SF313, SF314, SF315, SF316, SF317, SF321, SF322, SF323, SF324, SF335, SF326, SF325, SF326 , SF332, SF333, SF334, SF335, SF336, SF337 ... Star S0: processing target pixels S1, S2, S3, S4, S5, S6, S7, S8, S10, S11, S12, S13, S14, S15, S16, S17, S18, S, 20, S21, S22, S23, S24, S25, S26, S27, S28 ... comparison pixel PIX ... liquid crystal element

Claims (15)

  1.  映像信号に含まれる複数フレームの画素から形成される画素空間におけるノイズ除去の処理の対象となる処理対象画素と、前記画素空間において当該処理対象画素の周辺に位置する少なくとも3画素の比較画素とから比較画素群を構成し、当該比較画素群に属する複数の画素の画素値を比較し、前記処理対象画素の画素値が上位グループ、中位グループ及び下位グループのいずれに属するかの判定を行うノイズ判定部と、
     前記処理対象画素が前記上位グループに属する場合、当該処理対象画素の画素値から予め設定されるノイズ量を減算し、前記処理対象画素が前記下位グループに属する場合、当該処理対象画素の画素値から前記ノイズ量を加算し、前記処理対象画素が前記中位グループに属する場合、当該処理対象画素の画素値に対して前記ノイズ量を加減算しないノイズリダクション演算部と、
     を有することを特徴とする画像処理装置。
    A processing target pixel that is a target of noise removal processing in a pixel space formed from pixels of a plurality of frames included in a video signal, and at least three comparison pixels that are located around the processing target pixel in the pixel space Noise that constitutes a comparison pixel group, compares pixel values of a plurality of pixels belonging to the comparison pixel group, and determines whether the pixel value of the processing target pixel belongs to an upper group, a middle group, or a lower group A determination unit;
    When the pixel to be processed belongs to the upper group, a preset noise amount is subtracted from the pixel value of the pixel to be processed, and when the pixel to be processed belongs to the lower group, the pixel value of the pixel to be processed Adding the noise amount, and when the processing target pixel belongs to the middle group, a noise reduction calculation unit that does not add or subtract the noise amount with respect to the pixel value of the processing target pixel;
    An image processing apparatus comprising:
  2.  前記上位グループ、中位グループまたは下位グループそれぞれに含まれる画素数は、映像信号の種類に応じて設定されており、
     前記ノイズ判定部が前記比較画素群における少なくとも三つの画素の画素値を大きい順に順位付けし、前記上位グループ、前記中位グループ及び前記下位グループの各々に設定された前記画素数に応じて振り分けることを特徴とする請求項1に記載の画像処理装置。
    The number of pixels included in each of the upper group, middle group, or lower group is set according to the type of video signal,
    The noise determination unit ranks the pixel values of at least three pixels in the comparison pixel group in descending order, and distributes them according to the number of pixels set in each of the upper group, the middle group, and the lower group. The image processing apparatus according to claim 1.
  3.  前記映像信号の各フレームにおいて、前記処理対象画素と該処理対象画素の周辺の画素とを接続する線が、該フレームに含まれる輪郭線の一部を構成する場合における前記処理対象画素の周辺の画素を、前記比較画素として選択したことを特徴とする請求項1または請求項2に記載の画像処理装置。 In each frame of the video signal, a line connecting the processing target pixel and a pixel around the processing target pixel forms a part of a contour line included in the frame. The image processing apparatus according to claim 1, wherein a pixel is selected as the comparison pixel.
  4.  前記比較画素が、前記処理対象画素に対し、横方向、縦方向及び斜め方向に予め設定された位置の前記画素から選択されることを特徴とする請求項3に記載の画像処理装置。 The image processing apparatus according to claim 3, wherein the comparison pixel is selected from the pixels at positions preset in a horizontal direction, a vertical direction, and an oblique direction with respect to the processing target pixel.
  5.  前記比較画素が前記画素空間において前記処理対象画素を原点とし、x軸方向、y軸方向、前記原点を中心として時計方向に予め設定した角度回転させた軸方向、前記原点を中心として反時計方向に予め設定した角度回転させた軸方向にある画素から選択されていることを特徴とする請求項3または請求項4に記載の画像処理装置。 In the pixel space, the comparison pixel has the pixel to be processed as an origin, an x-axis direction, a y-axis direction, an axial direction rotated clockwise by a preset angle around the origin, and a counterclockwise direction around the origin The image processing apparatus according to claim 3, wherein the image processing apparatus is selected from pixels in an axial direction rotated by a preset angle.
  6.  前記画素空間が時間方向を加えた3次元空間の画素配列を含むことを特徴とする請求項1から請求項5のいずれか一項に記載の画像処理装置。 The image processing apparatus according to claim 1, wherein the pixel space includes a pixel array in a three-dimensional space with a time direction added.
  7.  前記画素空間が時間方向を加えた3次元空間の画素配列を含む場合において、1次のノイズ処理を行い、前記画素空間が前記処理対象画素の存在する2次元空間の画素配列の場合において、2次のノイズ処理を行うことを特徴とする請求項1から請求項5のいずれか一項に記載の画像処理装置。 When the pixel space includes a pixel array in a three-dimensional space with a time direction added, first-order noise processing is performed, and in the case where the pixel space is a pixel array in a two-dimensional space where the processing target pixel exists, 2 The image processing apparatus according to claim 1, wherein the following noise processing is performed.
  8.  前記予め設定されたノイズ量を算出するノイズ値検出回路をさらに有し、
     前記ノイズ値検出回路が、前記処理対象画素と、前記画素空間における当該処理対象画素の周辺における複数の他の画素との差分に基づいて、前記ノイズ量を算出することを特徴とする請求項1から請求項7のいずれか一項に記載の画像処理装置。
    A noise value detection circuit for calculating the preset noise amount;
    2. The noise amount detection circuit calculates the noise amount based on a difference between the processing target pixel and a plurality of other pixels around the processing target pixel in the pixel space. The image processing apparatus according to claim 7.
  9.  前記ノイズリダクション演算部が、前記中位グループの画素の画素値の平均値である中位グループ平均値を求め、前記処理対象画素が前記上位グループまたは前記下位グループに属する場合、当該処理対象画素の画素値から前記中位グループ平均値を減算し、
     前記ノイズ値検出回路は、前記ノイズリダクション演算部が減算した結果の絶対値を求め、前記前記2次元配列において選択された前記処理対象画素毎に得られた絶対値を前記2次元配列において平均化することにより前記ノイズ量を求める
     ことを特徴とする請求項8に記載の画像処理装置。
    The noise reduction calculation unit obtains a middle group average value that is an average value of pixel values of the pixels of the middle group, and when the processing target pixel belongs to the upper group or the lower group, Subtract the middle group average value from the pixel value,
    The noise value detection circuit obtains an absolute value as a result of subtraction by the noise reduction calculation unit, and averages the absolute value obtained for each of the processing target pixels selected in the two-dimensional array in the two-dimensional array. The image processing apparatus according to claim 8, wherein the amount of noise is obtained by doing so.
  10.  前記ノイズリダクション演算部が、前記処理対象画素が前記上位グループまたは前記下位グループに属する場合、当該処理対象画素の画素値が当該処理対象画素の画素値及び複数の前記比較画素の画素値のうちで最大値あるいは最小値であるか否かを判定し、最大値あるいは最小値である場合、当該処理対象画素の画素値と、各比較画素との差分の平均値である差分平均値を算出し
     前記ノイズ値検出回路は、前記ノイズリダクション演算部が算出した差分平均値の絶対値を算出し、該算出した差分平均値の絶対値を前記画素空間において平均化することにより前記ノイズ量を求める
     ことを特徴とする請求項8に記載の画像処理装置。
    When the processing target pixel belongs to the upper group or the lower group, the noise reduction calculation unit determines that the pixel value of the processing target pixel is the pixel value of the processing target pixel and the pixel values of the plurality of comparison pixels. It is determined whether or not it is the maximum value or the minimum value, and when it is the maximum value or the minimum value, a difference average value that is an average value of differences between the pixel value of the processing target pixel and each comparison pixel is calculated. The noise value detection circuit calculates an absolute value of the difference average value calculated by the noise reduction calculation unit, and calculates the noise amount by averaging the absolute value of the calculated difference average value in the pixel space. The image processing apparatus according to claim 8, wherein the apparatus is an image processing apparatus.
  11.  前記ノイズリダクション演算部が、前記処理対象画素が前記上位グループまたは前記下位グループに属する場合、当該処理対象画素の画素値から前記中位グループ平均値を減算することにより差分平均値を算出し、
     前記ノイズ値検出回路は、前記ノイズリダクション演算部が算出した差分平均値の絶対値を求め、前記差分平均値の絶対値を前記画素空間において平均化することにより前記ノイズ量を求める
     ことを特徴とする請求項8に記載の画像処理装置。
    When the processing target pixel belongs to the upper group or the lower group, the noise reduction calculation unit calculates a difference average value by subtracting the middle group average value from a pixel value of the processing target pixel,
    The noise value detection circuit calculates an absolute value of a difference average value calculated by the noise reduction calculation unit, and calculates the noise amount by averaging the absolute value of the difference average value in the pixel space. The image processing apparatus according to claim 8.
  12.  前記ノイズリダクション演算部が、前記処理対象画素が前記上位グループまたは前記下位グループに属する場合、当該処理対象画素の画素値が最大値あるいは最小値であるか否かを判定し、最大値あるいは最小値である場合、当該処理対象画素の画素値と、前記比較画素との差分の平均値である差分平均値を算出し、
     前記ノイズ値検出回路は、前記ノイズリダクション演算部が算出した差分平均値の絶対値を求め、1フレーム単位に当該差分平均値の数値毎にヒストグラムを生成し、該生成したヒストグラムの最頻値を前記ノイズ量とする
     ことを特徴とする請求項8に記載の画像処理装置。
    When the processing target pixel belongs to the upper group or the lower group, the noise reduction calculation unit determines whether the pixel value of the processing target pixel is the maximum value or the minimum value, and determines the maximum value or the minimum value. If it is, the difference average value which is the average value of the difference between the pixel value of the processing target pixel and the comparison pixel is calculated,
    The noise value detection circuit calculates an absolute value of the difference average value calculated by the noise reduction calculation unit, generates a histogram for each numerical value of the difference average value in units of one frame, and calculates a mode value of the generated histogram. The image processing apparatus according to claim 8, wherein the amount of noise is used.
  13.  ノイズ判定部が、映像信号に含まれる複数フレームの画素から形成される画素空間におけるノイズ除去の処理の対象となる処理対象画素と、前記画素空間において当該処理対象画素の周辺に位置する少なくとも3画素の比較画素とから比較画素群を構成し、当該比較画素群に属する複数の画素の画素値を比較し、前記処理対象画素の画素値が上位グループ、中位グループ及び下位グループのいずれに属するかの判定を行うノイズ判定過程と、
     ノイズリダクション演算部が、前記処理対象画素が前記上位グループに属する場合、当該処理対象画素の画素値から予め設定されるノイズ量を減算し、前記処理対象画素が前記下位グループに属する場合、当該処理対象画素の画素値から前記ノイズ量を加算し、前記処理対象画素が前記中位グループに属する場合、当該処理対象画素の画素値に対して前記ノイズ量を加減算しないノイズリダクション演算過程と、
     を含むことを特徴とする画像処理方法。
    The noise determination unit includes a processing target pixel that is a target of noise removal processing in a pixel space formed from pixels of a plurality of frames included in the video signal, and at least three pixels that are located around the processing target pixel in the pixel space A comparison pixel group is formed from the comparison pixels, pixel values of a plurality of pixels belonging to the comparison pixel group are compared, and whether the pixel value of the processing target pixel belongs to an upper group, a middle group, or a lower group A noise determination process for determining
    When the processing target pixel belongs to the upper group, the noise reduction calculation unit subtracts a preset noise amount from the pixel value of the processing target pixel, and when the processing target pixel belongs to the lower group, the processing Adding the noise amount from the pixel value of the target pixel, and when the processing target pixel belongs to the middle group, a noise reduction calculation process that does not add or subtract the noise amount to the pixel value of the processing target pixel;
    An image processing method comprising:
  14.  コンピュータに、
    映像信号に含まれる複数フレームの画素から形成される画素空間におけるノイズ除去の処理の対象となる処理対象画素と、前記画素空間において当該処理対象画素の周辺に位置する少なくとも3画素の比較画素とから比較画素群を構成し、当該比較画素群に属する複数の画素の画素値を比較し、前記処理対象画素の画素値が上位グループ、中位グループ及び下位グループのいずれに属するかの判定を行うノイズ判定ステップと、
     前記処理対象画素が前記上位グループに属する場合、当該処理対象画素の画素値から予め設定されるノイズ量を減算し、前記処理対象画素が前記下位グループに属する場合、当該処理対象画素の画素値から前記ノイズ量を加算し、前記処理対象画素が前記中位グループに属する場合、当該処理対象画素の画素値に対して前記ノイズ量を加減算しないノイズリダクション演算ステップと、
     を実行させるための画像処理プログラム。
    On the computer,
    A processing target pixel that is a target of noise removal processing in a pixel space formed from pixels of a plurality of frames included in a video signal, and at least three comparison pixels that are located around the processing target pixel in the pixel space Noise that constitutes a comparison pixel group, compares pixel values of a plurality of pixels belonging to the comparison pixel group, and determines whether the pixel value of the processing target pixel belongs to an upper group, a middle group, or a lower group A determination step;
    When the pixel to be processed belongs to the upper group, a preset noise amount is subtracted from the pixel value of the pixel to be processed, and when the pixel to be processed belongs to the lower group, the pixel value of the pixel to be processed When the noise amount is added and the processing target pixel belongs to the middle group, a noise reduction calculation step that does not add or subtract the noise amount with respect to the pixel value of the processing target pixel;
    An image processing program for executing
  15.  請求項1に記載の画像処理装置を有し、映像信号のノイズ除去を行った後に映像表示を行うことを特徴とする表示装置。 A display device comprising: the image processing device according to claim 1, wherein video display is performed after noise removal from the video signal.
PCT/JP2011/074921 2010-11-02 2011-10-28 Image processing apparatus, image processing method, and display apparatus WO2012060290A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
JP2010246676 2010-11-02
JP2010-246676 2010-11-02

Publications (1)

Publication Number Publication Date
WO2012060290A1 true WO2012060290A1 (en) 2012-05-10

Family

ID=46024404

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/JP2011/074921 WO2012060290A1 (en) 2010-11-02 2011-10-28 Image processing apparatus, image processing method, and display apparatus

Country Status (1)

Country Link
WO (1) WO2012060290A1 (en)

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH07250264A (en) * 1994-03-10 1995-09-26 Nippon Hoso Kyokai <Nhk> Noise reducing circuit
JPH08172368A (en) * 1994-12-16 1996-07-02 Hitachi Ltd Sn ratio improving means
JPH09163185A (en) * 1995-12-12 1997-06-20 Hitachi Ltd S/n improving device
JPH1084498A (en) * 1996-09-09 1998-03-31 Nec Corp Noise reduction circuit
JPH10248027A (en) * 1997-03-04 1998-09-14 Hitachi Ltd S/n improving device
JP2000022992A (en) * 1998-07-02 2000-01-21 Hitachi Ltd Noise reducing signal processor
JP2008205737A (en) * 2007-02-19 2008-09-04 Olympus Corp Imaging system, image processing program, and image processing method

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH07250264A (en) * 1994-03-10 1995-09-26 Nippon Hoso Kyokai <Nhk> Noise reducing circuit
JPH08172368A (en) * 1994-12-16 1996-07-02 Hitachi Ltd Sn ratio improving means
JPH09163185A (en) * 1995-12-12 1997-06-20 Hitachi Ltd S/n improving device
JPH1084498A (en) * 1996-09-09 1998-03-31 Nec Corp Noise reduction circuit
JPH10248027A (en) * 1997-03-04 1998-09-14 Hitachi Ltd S/n improving device
JP2000022992A (en) * 1998-07-02 2000-01-21 Hitachi Ltd Noise reducing signal processor
JP2008205737A (en) * 2007-02-19 2008-09-04 Olympus Corp Imaging system, image processing program, and image processing method

Similar Documents

Publication Publication Date Title
KR101138852B1 (en) Smart clipper for mobile displays
US6717622B2 (en) System and method for scalable resolution enhancement of a video image
RU2413384C2 (en) Device of image processing and method of image processing
US8280190B2 (en) Image display device, image processing circuit, and image processing method
JP2008092462A (en) Contour correcting method, image processing apparatus and display device
EP2317757B1 (en) Image processing apparatus, image processing method, program and storage medium
US8379997B2 (en) Image signal processing device
KR20070040393A (en) Video processor comprising a sharpness enhancer
US8331659B2 (en) Method and apparatus for reduced complexity video processing via special chroma handling
US20100201701A1 (en) Image processor, image processing method, display device, program and integrated circuit
JP2003069859A (en) Moving image processing adapting to motion
WO2012060290A1 (en) Image processing apparatus, image processing method, and display apparatus
JP4843478B2 (en) Image processing apparatus and image processing method
WO2012161213A1 (en) Image processing device and image processing method
US20100104212A1 (en) Contour correcting device, contour correcting method and video display device
US6163308A (en) Method and apparatus for minimizing visual artifacts caused by the pixel display of a video image
US8270773B2 (en) Image processing apparatus and image processing method
US8289348B2 (en) Image signal processing device
TWI390958B (en) Video filter and video processor and processing method using thereof
JP2004361497A (en) Image display device, image display method, and image display program
JP2008148296A (en) Image processing circuit, semiconductor device, and image processing apparatus
WO2012017773A1 (en) Signal processing apparatus, signal processing method, signal processing program, and display apparatus
CN100463030C (en) Plasma display device and image processing method thereof
JP5193976B2 (en) Video processing apparatus and video display apparatus
WO2012124516A1 (en) Noise reduction processing device, and display device

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 11837940

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 11837940

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: JP