WO2011134144A1 - Image downscaling method and apparatus - Google Patents

Image downscaling method and apparatus Download PDF

Info

Publication number
WO2011134144A1
WO2011134144A1 PCT/CN2010/072270 CN2010072270W WO2011134144A1 WO 2011134144 A1 WO2011134144 A1 WO 2011134144A1 CN 2010072270 W CN2010072270 W CN 2010072270W WO 2011134144 A1 WO2011134144 A1 WO 2011134144A1
Authority
WO
WIPO (PCT)
Prior art keywords
dot matrix
processing device
media processing
nxn
pixel
Prior art date
Application number
PCT/CN2010/072270
Other languages
French (fr)
Chinese (zh)
Inventor
李琛
刘俊秀
石岭
Original Assignee
深圳艾科创新微电子有限公司
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 深圳艾科创新微电子有限公司 filed Critical 深圳艾科创新微电子有限公司
Priority to PCT/CN2010/072270 priority Critical patent/WO2011134144A1/en
Priority to CN201080061138.2A priority patent/CN102893592B/en
Publication of WO2011134144A1 publication Critical patent/WO2011134144A1/en

Links

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N1/00—Scanning, transmission or reproduction of documents or the like, e.g. facsimile transmission; Details thereof
    • H04N1/387—Composing, repositioning or otherwise geometrically modifying originals
    • H04N1/393—Enlarging or reducing
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T3/00—Geometric image transformations in the plane of the image
    • G06T3/40—Scaling of whole images or parts thereof, e.g. expanding or contracting
    • G06T3/4023—Scaling of whole images or parts thereof, e.g. expanding or contracting based on decimating pixels or lines of pixels; based on inserting pixels or lines of pixels

Definitions

  • the present invention relates to image processing techniques, and more particularly to an image reduction method and apparatus.
  • Image scaling techniques are commonly used in a variety of media processing devices.
  • the media processing device When the image signal enters the media processing device, the media processing device provides the image signal to the display device. Since the resolution of the display device is often different from the resolution of the signal source, the media processing device often needs to use an image scaling technique to scale the image signal, and then provide the scaled image signal to the display device.
  • the image includes a video image and a picture.
  • the more common high-definition image resolution is 1440X1080. If you need to display a high-definition image with a resolution of 1440X1080 on a display device with a resolution of 640X480, you need to reduce the image with a resolution of 1440X1080 by 2.25 (horizontal direction) X2. After 25 (vertical) times, it can be displayed on a display device with a resolution of 640X480.
  • the decoder performs a discrete cosine transform (DCT, Discrete Cosine Transform) decoding on the compressed image signal, and the compressed image signal is, for example, an image signal in the format of the Moving Picture Experts Group (MPEG).
  • DCT discrete cosine transform
  • MPEG Moving Picture Experts Group
  • the image signal of the Joint Picture Expert Group (JPEG) format the decoder inputs the image signal decoded by the DCT decoding technology to the media processing device, and the media processing device performs the image signal decoded by the DCT decoding technology. Processing, and inputting the processed image signal to the display device display.
  • JPEG Joint Picture Expert Group
  • the decoder generally outputs the image signal decoded by the DCT decoding technology in units of 8 ⁇ 8 dot matrix. After the media processing device obtains the image signal in units of 8 ⁇ 8 dot matrix, it is generally necessary to buffer a plurality of 8 ⁇ 8 dot matrix, and then, Then reduce the processing in units of actions.
  • the media processing device after the media processing device obtains the image signal in units of 8 ⁇ 8 dot matrix, before the image signal is reduced, a plurality of 8 ⁇ 8 dot matrix needs to be cached, so that the media processing device needs to cache resources to cache. Multiple 8X8 dot matrix, such cache resources for media processing
  • the device is not a small resource overhead.
  • the traditional method mostly uses multi-phase interpolation, which is to multiply and add multiple samples to obtain the output of one pixel, and also needs anti-aliasing filtering.
  • the reduction processing method is more complicated.
  • the present invention provides an image reduction method and apparatus for saving cache resources of a media processing device when performing image reduction processing on an image signal in units of 8 ⁇ 8 dots.
  • the invention provides an image reduction method, which is suitable for an image preview mode, and is applied to an image processing system including a decoder and a media processing device, and the media processing device reduces the image signal output by the decoder in units of 8 ⁇ 8 dot matrix.
  • the image signal in the 8 ⁇ 8 dot matrix is an image signal obtained by the decoder after DCT decoding, and the method includes: the media processing device obtains an image signal in units of 8 ⁇ 8 dot matrix; the media processing device is currently obtained A pixel is selected in the 8X8 dot matrix for participating in a process of reducing the image signal in units of 8 ⁇ 8 dot matrix by N times, wherein N is a positive integer; the pixel selected by the media processing device from the currently obtained 8 ⁇ 8 dot matrix Performing an operation; if N is less than 8, the pixel corresponding to the NXN lattice is obtained; if N is greater than 8, an intermediate result corresponding to the NXN lattice is obtained; wherein, for each NXN lattice, one pixel is obtained, if N is greater than 8.
  • the pixel points obtained for each NXN lattice are obtained by computing multiple intermediate results.
  • the present invention also provides a media processing device suitable for an image preview mode, applied to an image processing system including a decoder and a media processing device, wherein the media processing device performs an image signal output by the decoder in an 8 ⁇ 8 dot matrix unit.
  • the image signal in the unit of 8 ⁇ 8 dot matrix is an image signal obtained by the decoder after DCT decoding
  • the media processing device includes: an obtaining unit, configured to obtain an image signal in units of 8 ⁇ 8 dot matrix; a unit, configured to select a pixel from the currently obtained 8 ⁇ 8 dot matrix, and the selected pixel is used to participate in a process of reducing an image signal of 8 ⁇ 8 dot matrix by N times, wherein N is a positive integer; Calculating the pixel points selected from the currently obtained 8 ⁇ 8 dot matrix; if N is less than 8, the pixel points corresponding to the NXN lattice are obtained; if N is greater than 8, the intermediate result corresponding to the NXN lattice is obtained;
  • N
  • the image signal for participating in the 8 ⁇ 8 dot matrix is selected from the currently obtained 8 ⁇ 8 dot matrix to be reduced by N.
  • Double the processed pixel points calculate the selected pixel points, output the pixel points or intermediate results corresponding to the NXN dot matrix, and do not need to cache multiple 8X8 dot matrixes, then perform reduction processing, saving multiple 8X8 dot matrix for buffering And the cache resources needed.
  • FIG. 1 is a schematic diagram showing a logical structure of an image processing system to which the present invention is applied;
  • Figure 2 is a schematic diagram of an 8X8 dot matrix
  • FIG. 3 is a flow chart of an image reduction method of the present invention.
  • Figure 4A is a diagram showing the division of an 8X8 dot matrix in the case where the reduction factor is 2 in one embodiment of the present invention
  • 4B is a schematic diagram of a basic template in the case where the reduction factor is 2 in one embodiment of the present invention.
  • Figure 5A is a diagram showing the division of an 8X8 dot matrix in the case where the reduction factor is 4 in one embodiment of the present invention
  • Figure 5B is a schematic diagram of a basic template in the case where the reduction factor is 4 in one embodiment of the present invention.
  • Figure 6A is a diagram showing the division of 3X3 8X8 dot matrix in the case where the reduction factor is 3 in one embodiment of the present invention
  • 6B and 6C are schematic diagrams showing a basic template in the case where the reduction factor is 3 in one embodiment of the present invention.
  • Figure 7 is a diagram showing the composition of a 16X16 dot matrix in the case where the reduction factor is 16 in one embodiment of the present invention.
  • FIG. 8 is a schematic diagram of a plurality of 8 ⁇ 8 dot matrix groups formed into a plurality of 12 ⁇ 12 dot matrix in the case where the reduction factor is 12 in one embodiment of the present invention
  • FIG. 9 is a flow chart showing an implementation of the image reduction method of the present invention.
  • FIG. 10 is a schematic diagram showing the logical structure of a media processing device of the present invention. detailed description
  • the image processing system includes a decoder, a media processing device, and a display device.
  • the decoder outputs an image signal in units of 8 ⁇ 8 dot matrix
  • the media processing device performs a reduction process on the image signal in units of 8 ⁇ 8 dot matrix, and outputs the pixel point to the display device.
  • an 8 X 8 dot matrix consists of 64 pixels, distributed in 8 rows and 8 columns, each row consisting of 8 pixels, each column consisting of 8 pixels, and any pixel is in 8X8 dot matrix. The position in this is determined by the row and column of this pixel in the 8X8 dot matrix.
  • the rows in the 8X8 dot matrix are represented by i (i is a positive integer not greater than 8), and the column is represented by j (j is a positive integer not greater than 8), then any pixel in the 8X8 dot matrix can be used.
  • a (i, j) to indicate.
  • the media processing device when receiving the image signal in units of 8 ⁇ 8 dot matrix, receives the signal in units of 8 ⁇ 8 dots, that is, after receiving the first 8 ⁇ 8 dot matrix, and then receives the second image.
  • 8 X8 dot matrix after receiving all the pixels of the first 8 rows, then receive the first 8x8 dot matrix of the second 8 rows, the 2nd 8x8 dot matrix, until the second 8 rows are accepted After all the pixels, the first 8 X 8 dot matrix, the 2nd 8 X 8 dot matrix of the third 8 rows, and so on, until all 8X8 dot matrixes are received.
  • the process of reducing the image signal by the unit of 8 ⁇ 8 dot matrix by N times is to reduce the image signal by N times in the horizontal direction and the vertical direction, that is, to reduce the image signal in units of 8 ⁇ 8 dot matrix.
  • N (horizontal direction) XN (vertical direction) times N-folding the image signal in units of 8 ⁇ 8 dot matrix is just a convenient way to describe it.
  • N-folding the image signal can be considered as N-folding the image, or it can be considered to be N times smaller for each 8X8 dot matrix that makes up the image, where N is a positive integer.
  • one pixel point is output correspondingly for each NXN dot matrix.
  • the basic template mentioned in the following embodiments of the present invention is a template used by the media processing device to select pixels, and the template defines pixel points selected from the dot matrix.
  • the design of the template is not unique, or It is said that which pixel points are selected from the dot matrix is not unique, and can be designed by a person skilled in the art according to actual needs.
  • the information of one pixel can be represented by multiple components, such as R,
  • each component of the pixel should be operated separately.
  • one pixel corresponds to three components, namely component 1, component 2 and component 3.
  • the component 1 of the pixel should be calculated to obtain the component 1 of the output pixel.
  • the component 2 of the pixel is calculated to obtain the component 2 of the output pixel, and the component 3 of the pixel is calculated to obtain the component 3 of the output pixel, so that three components of the output pixel are obtained, and thus The pixel point of the output.
  • All of the embodiments of the present invention are applicable to the image preview mode. All embodiments of the present invention can be applied to an image processing system including a decoder and a media processing device, and the media processing device performs a reduction process on the image signal of the 8 ⁇ 8 dot matrix output by the decoder to 8 ⁇ 8 points.
  • the image signal of the array is the image signal obtained by the decoder after DCT decoding.
  • the media processing device obtains an image signal in units of 8 ⁇ 8 dots.
  • the media processing device selects a pixel from the currently obtained 8 ⁇ 8 dot matrix for participating in a process of reducing an image signal of 8 ⁇ 8 dot matrix by N times, wherein N is a positive integer.
  • S303 The media processing device performs operations on the pixels selected from the currently obtained 8 ⁇ 8 dot matrix; if N is less than 8, the pixel points corresponding to the NXN lattice are obtained; if N is greater than 8, the corresponding NXN lattice is obtained. Intermediate result; wherein, for each NXN lattice, one pixel is obtained. If N is greater than 8, the pixel obtained for each NXN lattice is calculated by using multiple intermediate results.
  • the media processing device When the media processing device operates on the pixels selected from the currently obtained 8 X 8 dot matrix, the selected pixel points can be averaged.
  • the media processing device can store the intermediate result. After obtaining all the intermediate results of the corresponding NXN lattice, the media processing device calculates all the intermediate results and obtains the pixels corresponding to the NXN lattice.
  • the media processing device When the media processing device operates on the pixels selected from the currently obtained 8 ⁇ 8 dot matrix, The summation is performed on the pixels selected from the currently obtained 8 ⁇ 8 dot matrix.
  • the intermediate result of the corresponding NXN lattice is the result of the summation operation performed by the media processing device on the pixel points selected from the currently obtained 8 ⁇ 8 dot matrix.
  • the result of all the summation operations can be averaged to obtain the pixel points corresponding to the NXN lattice. Specifically, if N is a divisor of 8, then an 8 ⁇ 8 dot matrix can be divided into (8/N) (horizontal direction) X (8/N) (vertical direction) NXN lattices.
  • the media processing device may respectively select pixels from each of the currently obtained 8 ⁇ 8 lattices, and respectively select each of the NXN lattices from the currently obtained 8 ⁇ 8 dot matrix.
  • the pixels are operated to obtain pixels corresponding to each of the currently obtained 8 ⁇ 8 lattices.
  • an 8X8 dot matrix can be divided into 4X4 2X2 dot matrixes, each of which outputs one pixel point, and the output pixel points are recorded as 0 (i, j).
  • the basic template is as shown in FIG. 4B.
  • all the pixels in each 2 ⁇ 2 dot matrix participate in the operation of outputting pixel points, and the operation manner is to take the values of all the pixel points in the basic template.
  • the average value that is, all the pixels are selected from the 2X2 dot matrix for the mean operation.
  • 4 X 4 pixel points of an 8X8 dot matrix output can be obtained in sequence, as follows:
  • 0(1, 1) [A(l, 1) + A(1, 2) +A(2, 1)+A(2, 2)]/ 4
  • 0(1, 2) [A(l, 3) + A(l, 4)+A(2, 3)+A(2, 4) ]/4
  • an 8 ⁇ 8 dot matrix can be divided into 2 ⁇ 2 4 ⁇ 4 dot matrixes, each of which outputs one pixel point, and the output pixel points are recorded as 0 (i, j).
  • the basic template is as shown in FIG. 5B, and the shadow pixel points in each 4 ⁇ 4 dot matrix participate in the operation of outputting pixel points in such a manner as to take the average value of the values of the shaded pixel points in the basic template.
  • 0(1, 2) [A(l, 5) + A(l, 7) +A (3, 5)+A(3, 7)]/ 4
  • 0(2, 1) [A(5, 1) + A(5, 3)+A(7, 1)+A(7, 3)]/ 4
  • an 8X8 lattice can be divided into at least one NXN lattice and a plurality of boundary pixels, and boundary pixels in an 8 ⁇ 8 lattice can be associated with adjacent 8 ⁇ 8 lattices.
  • the boundary pixels constitute an NXN lattice, wherein the boundary pixel here refers to a pixel in the 8 ⁇ 8 dot matrix other than the pixel included in the NXN dot matrix, or 8 ⁇ 8 dot matrix and adjacent 8 X 8
  • the pixels in the dot matrix participate in the pixel points that make up the NXN dot matrix.
  • the media processing device can each from each of the currently acquired 8X8 lattices.
  • the media processing device can Selecting pixel points from boundary pixels used to form the NXN lattice in the currently obtained 8 ⁇ 8 dot matrix, and selecting from the currently selected 8 ⁇ 8 dot matrix used to form the boundary pixels of the NXN lattice The pixel is operated to obtain a pixel corresponding to the N XN dot matrix.
  • the media processing device needs to reduce the received image signal by a factor of three.
  • an 8 X 8 dot matrix needs to output 2.67 X 2.67 pixels, but in practice, there are no non-integer number of pixels.
  • the least common multiple of 3 and 8 is 24, so that 3X3 8 X8 dot matrix outputs 8X8 pixels.
  • 3X3 8X8 dots can be divided into 8X8 3X3 dot matrix, each 3X3 dot matrix outputs one pixel, and the output pixel is recorded as 0 a").
  • the pixels in the first 8 ⁇ 8 dot matrix, the pixels in the second 8 ⁇ 8 dot matrix, and the pixels in the third 8 ⁇ 8 dot matrix belong to the same 8-row pixel point.
  • the pixel in the 4th 8X8 dot matrix, the pixel in the 5th 8X8 dot matrix, and the pixel in the 6th 8X8 dot matrix belong to the same 8 rows of pixels, and the pixels in the 7th 8X8 dot matrix Point, 8th 8
  • the pixel points in the X 8 dot matrix and the pixel dots in the 9th 8 X 8 dot matrix belong to the same 8 rows of pixel points.
  • the adjacent 8 X 8 lattices of the first 8 X 8 dot matrix are the 2nd 8 X 8 dot matrix and the 4th 8 X 8 dot matrix respectively, and the 5X 8 8 dot matrix adjacent 8X8 dot matrix are respectively 1st 8X8 dot matrix, 2nd 8X8 dot matrix, 3rd 8X8 dot matrix, 4th 8X8 dot matrix, 6th 8X8 dot matrix, 7th 8X8 dot matrix, 8th 8X8 dot matrix and 9 8X8 dot matrixes, and other 8X8 dot matrix adjacent 8X8 dot matrix can be determined accordingly, and will not be described here.
  • the media processing device after receiving the first 8 ⁇ 8 dot matrix, the second 8 ⁇ 8 dot matrix, and the third 8 ⁇ 8 dot matrix respectively. It is also necessary to receive the other 8X8 lattices in the same row in turn, and then receive the 4th 8X8 dot matrix, the 5th 8X8 dot matrix, and the 6th 8X8 dot matrix in turn, in the same way, after accepting the 4th respectively.
  • 8X8 dot matrix 5th 8X8 dot matrix and 6th 8X8 dot matrix, you need to receive the other 8X8 dot matrix in the same row as the 3 8X8 dot matrix in turn, and then receive the 7th 8X8 dot matrix in turn. , 8th 8X8 dot matrix and 9th 8X8 dot matrix.
  • the basic templates of the first 8 ⁇ 8 dot matrix are respectively shown in FIG. 6B and FIG. 6C, and the first 8 ⁇ 8 dot matrix outputs 3 ⁇ 3 pixel points, and the 3 ⁇ 3 pixel points are:
  • the pixel points in the seventh and eighth columns of the first 8 ⁇ 8 dot matrix and the pixel points in the first column of the second 8 ⁇ 8 dot matrix can be combined into three 3 ⁇ 3 dot matrixes, due to the first 8 ⁇ 8
  • the number of pixels in the 7th and 8th columns of the dot matrix is large, so from the 7th and 8th columns of the 1st 8X8 dot matrix
  • the pixel is selected among the pixels to participate in the operation of outputting the pixel.
  • the pixels in the 7th and 8th rows of the 1st 8X8 dot matrix and the pixels in the 1st row of the 4th 8X8 dot matrix can be combined into 3 3X3 dot matrixes, due to the first 8X8 dot matrix.
  • the number of pixels is large. Therefore, the pixel point is selected from the pixels in the seventh and eighth rows of the first 8 ⁇ 8 dot matrix to participate in the operation of outputting the pixel point.
  • the basic template of the second 8X8 dot matrix is as shown in Fig. 6B. Since the pixels in the 7th and 8th columns of the 1st 8X8 dot matrix have already output 3 pixels, the pixel in the 1st column of the 2nd 8X 8 dot matrix is not in the 2nd 8X8 dot matrix. Participate in the operation of outputting pixel points. Similarly, the pixel in the eighth column of the second 8 ⁇ 8 dot matrix and the pixel in the first and second columns of the third 8 ⁇ 8 dot matrix can also be combined into three 3 ⁇ 3 dot matrix, due to the third 8 ⁇ 8 dot. The number of pixels in the first and second columns of the array is large. Therefore, the pixels in the first and second columns of the third 8 ⁇ 8 dot matrix output 3 pixels, and the eighth column of the second 8 ⁇ 8 dot matrix. The pixel points do not participate in the operation of outputting pixel points in the second 8 ⁇ 8 dot matrix.
  • the second 8X8 dot matrix output 2X3 pixels, 2X3 pixels are as follows:
  • the basic templates of the fourth 8 X 8 dot matrix are shown in Figs. 6B and 6C, respectively. Since the pixels in the 7th and 8th rows of the 1st 8X8 dot matrix have already output 3 pixels, the pixel in the 1st row of the 4th 8 ⁇ 8 dot matrix is in the 8 ⁇ 8 dot matrix. Does not participate in the operation of outputting pixel points. Similarly, the pixels in the 8th row of the 8X8 dot matrix and the pixels in the 1st and 2nd rows of the 7th 8X8 dot matrix can also be combined into three 3X3 dot matrixes, due to the 7th 8X8 dot matrix In the 7th and 8th rows, the number of pixels is large.
  • the pixels in the 7th and 8th rows of the 7th 8X8 dot matrix output 3 pixels, and the pixels in the 8th row of the 4th 8X8 dot matrix.
  • the point does not participate in the operation of outputting pixel points in the 8X8 dot matrix.
  • the values of the pixels of the other 8 X 8 dot matrix output are the same as those of the above calculation method, and will not be described here.
  • an 8X8 dot matrix needs to output 1.6X1.6 pixels, but in practical applications, there are no non-integer number of pixels.
  • the least common multiple of 5 and 8 is 40, so that 5X5 8 X 8 dot matrix outputs 8 X 8 pixels.
  • 5X5 8X8 dots can be divided into 8X8 5X5 dots, and each 5X5 dot matrix outputs one pixel.
  • the media processing device can get an intermediate result, and the intermediate result corresponds to the N X N dot matrix composed of the currently acquired 8X8 dot matrix.
  • the intermediate processing result can be stored after the media processing device gets an intermediate result corresponding to the NXN lattice. After obtaining all the intermediate results of the corresponding NXN lattice, the media processing device can calculate all the intermediate results to obtain the pixels corresponding to the NXN lattice. For example, the media processing device may perform a summation operation on the pixels selected from the currently obtained 8 ⁇ 8 dot matrix, and the result of the summation operation is an intermediate result, and the media processing device may perform an average operation on the results of all the summation operations.
  • the denominator used in the mean operation is the number of pixels participating in the operation, that is, the currently acquired 8 ⁇ 8 dot matrix participates in all the pixels of the NXN lattice participating in the operation. Number.
  • each 2 X2 8X8 dot matrix outputs one pixel.
  • the media processing device may add the values of the pixels participating in the operation of the output pixel in the first 8 ⁇ 8 dot matrix, and buffer the result.
  • the media processing device may add the values of the pixels participating in the operation of the output pixel in the first 8 ⁇ 8 dot matrix, and buffer the result.
  • the second 8 ⁇ 8 dot matrix The values of the pixels participating in the operation of the output pixel in the second 8 ⁇ 8 dot matrix can be added, and the result can be cached.
  • the third 8 ⁇ 8 dot matrix is obtained, the third 8 ⁇ 8 dot matrix can participate in the output pixel point.
  • the values of the calculated pixel points are added, and the result of the buffering is added.
  • the values of the pixel points of the operation of the output pixel point in the fourth 8 ⁇ 8 dot matrix can be added to obtain a result, and finally The average of the above four results is obtained to obtain the value of the output pixel.
  • the denominator used in the average calculation is the first 8 ⁇ 8 dot matrix, the second 8 ⁇ 8 dot matrix, and the third 8 ⁇ 8 dot matrix.
  • the 4th 8X8 The number of pixels in the lattice that participate in the operation.
  • an NXN lattice consists of at least one 8 ⁇ 8 dot matrix and boundary pixels in an adjacent 8 ⁇ 8 dot matrix, where the pixel points are adjacent to the 8 ⁇ 8 dot matrix.
  • the pixels in the 8 X 8 dot matrix participate in the pixels that make up the NXN dot matrix.
  • the media processing device can select pixels from all the pixels in the currently obtained 8 ⁇ 8 dot matrix, and the 8X8 dot matrix obtained from the current one.
  • the pixel selected from all the pixels in the operation is operated to obtain an intermediate result corresponding to the NXN lattice composed of the currently obtained 8 ⁇ 8 lattice.
  • the media processing device selects pixel points from the boundary pixel points in the currently obtained 8 ⁇ 8 dot matrix, and selects from the currently obtained 8 ⁇ 8 dot matrix.
  • the pixels selected in the boundary pixels are operated to obtain an intermediate result corresponding to the NXN lattice composed of the currently obtained 8 ⁇ 8 lattice.
  • the media processing device is getting the corresponding NXN point
  • intermediate results can be stored.
  • the media processing device can calculate all the intermediate results to obtain the pixel points corresponding to the NXN lattice. For example, the media processing device may perform a summation operation on the pixels selected from the currently obtained 8 ⁇ 8 dot matrix, and the result of the summation operation is an intermediate result, and the media processing device may perform an average operation on the results of all the summation operations.
  • the denominator used in the mean operation is the number of pixels participating in the operation, that is, all the pixels participating in the NXN lattice that the currently obtained 8 ⁇ 8 lattice participates in. Number.
  • the media processing device needs to reduce the received image signal by 12 times.
  • an 8 ⁇ 8 dot matrix needs to output 0.67 ⁇ 0.67 pixels, but in practical applications, there is no non-integer number.
  • 3X3 8X8 dots can be divided into 2X2 12X12 lattices, and each 12X12 dot matrix outputs one pixel. As shown in Fig.
  • the first 12X12 dot matrix after division is composed of the first 8 ⁇ 8 dot matrix, the second 8 ⁇ 8 dot matrix 1-4 column, the fourth 8 ⁇ 8 dot matrix row 1-4 and the
  • the pixels in the 1-4th column of the 1-4th row of the 5 8X8 dot matrix are composed of the same, and the composition principle of the other 12X 12 dot matrix is the same.
  • the values of the pixels of other 12X12 dot matrix outputs can also be obtained by operation.
  • the media processing device receives the 8 ⁇ 8 dot matrix shown in FIG. 8 in sequence, therefore, if the first 8 ⁇ 8 dot matrix, the second 8 ⁇ 8 dot matrix, the first 1-4 column, the first In the 1-4th row of the 4th 8X8 dot matrix and the 1-4th row of the 1-4th row of the 5X8 dot matrix, there are pixels in the 1-4th column of the 8X8 dot matrix that need to participate in the operation of the output pixel point, then the media processing device obtains the After an 8X8 dot matrix, the values of the pixels participating in the operation of the output pixel in the first 8 ⁇ 8 dot matrix can be added, and the result can be cached.
  • the second 8 ⁇ 8 dot can be obtained.
  • the values of the pixels participating in the operation of the first 12X12 dot matrix output pixel are added, and the result is cached.
  • the fourth 8 X8 dot matrix can participate in the first 12X.
  • the value of the pixel point of the operation of the 12-dot output pixel is added, and the result is cached.
  • the 5th 8X8 dot matrix the 5th bit can be obtained.
  • the values of the pixels participating in the operation of the first 12 ⁇ 12 dot matrix output pixel in the 8 ⁇ 8 dot matrix are added to obtain a result. Finally, the above four results are averaged to obtain the first 12 ⁇ 12 dot matrix output.
  • the value of the pixel, where the denominator used in the average operation is the pixel of the first 8 ⁇ 8 dot matrix, the second 8 ⁇ 8 dot matrix, the 4th 8 ⁇ 8 dot matrix, and the 5th 8 ⁇ 8 dot matrix.
  • the values of the other 12X12 dot matrix output pixel points are handled in the same manner, and will not be described here.
  • the media processing device may also obtain pixel points corresponding to the N X N dot matrix by other means. For example, the media processing device may select pixel points only from the 8 ⁇ 8 dot matrix that participates in the largest number of pixels that make up the NXN dot matrix, and calculate the selected pixel points to obtain pixel points corresponding to the NXN dot matrix.
  • the 8X8 dot matrix that participates in the largest number of pixels that make up the NXN lattice represents the NXN dot matrix.
  • N 12 as an example.
  • the first 12X12 dot matrix after division is composed of the first 8X8 dot matrix, the second 8 ⁇ 8 dot matrix, the 1-4th column, and the fourth 8X8.
  • the pixels in the 1-4th row of the dot matrix and the 1-4th row in the 1-4th row of the 5th 8X8 dot matrix are composed of the pixel points corresponding to the 12X12 dot matrix, due to the first 8X8 dot.
  • the array is the 8X8 dot matrix that constitutes the largest number of pixels in the first 12X12 dot matrix. Therefore, pixels can be selected only from the first 8 ⁇ 8 dot matrix, and the selected pixel points are calculated to obtain the corresponding first pixel.
  • the pixel of the 12X12 dot matrix may include:
  • S902 Determine whether N is less than 8, if yes, then go to S903, otherwise, go to S908;
  • S903 Determine whether N is a divisor of 8, if yes, then go to S904, otherwise, go to S906;
  • S904 select the basic template corresponding to the NXN lattice, and then go to S905;
  • S905 Select, according to the selected basic template, pixel points from each of the NXN lattices in the 8 ⁇ 8 dot matrix, and perform operations on the selected pixel points respectively, respectively obtaining each of the 8 ⁇ 8 dot matrixes.
  • the pixel corresponding to the NXN dot matrix turn to S913;
  • S906 Select a basic template corresponding to the NXN lattice and a basic template corresponding to the boundary pixel, and then go to S907;
  • S907 Select, according to the selected basic template, pixel points from each of the NXN dot matrix and the boundary pixel in the 8 ⁇ 8 dot matrix, and perform operations on the selected pixel points respectively, respectively obtaining each NXN in the 8 ⁇ 8 dot matrix.
  • S908 Determine whether N is a multiple of 8, if yes, then go to S909, otherwise, go to S911;
  • S909 select the basic template corresponding to 8 X 8 dot matrix, and turn to S910;
  • S910 Select, according to the selected basic template, pixel points from each 8 ⁇ 8 dot matrix participating in the NXN dot matrix, and perform operations on the selected pixel points respectively to obtain multiple intermediate results, where the components participating in the NXN dot matrix are formed.
  • Each 8X8 dot matrix corresponds to an intermediate result, after which all the intermediate results are calculated to obtain the pixel points corresponding to the NXN dot matrix, and the process proceeds to S913;
  • S911 selecting a basic template corresponding to all pixel points in the 8 ⁇ 8 dot matrix that constitutes the N X N dot matrix and boundary pixel points in the 8 ⁇ 8 dot matrix, and then turning to S912;
  • S912 Select, according to the selected basic template, pixel points from each 8 ⁇ 8 dot matrix participating in the NXN dot matrix, and perform operations on the selected pixel points respectively to obtain a plurality of intermediate results, wherein, the components participating in the NXN dot matrix are formed.
  • Each 8X8 dot matrix corresponds to an intermediate result, after which all the intermediate results are calculated to obtain the pixel points corresponding to the NXN dot matrix, and the process proceeds to S913;
  • the present invention also provides a media processing device.
  • the media processing device includes: an obtaining unit 1001, configured to obtain an image signal in units of 8 ⁇ 8 dot matrix; and a selecting unit 1002, configured to select a pixel point from the currently obtained 8 ⁇ 8 dot matrix, and selected The pixel is used to participate in a process of reducing the image signal in units of 8 ⁇ 8 dot matrix by N times, wherein N is a positive integer; and the operation unit 1003 is configured to perform operation on the pixel selected from the currently obtained 8 ⁇ 8 dot matrix; If N is less than 8, the pixel corresponding to the NXN lattice is obtained; if N is greater than 8, an intermediate result corresponding to the NXN lattice is obtained; wherein, for each NXN lattice, one pixel is obtained, and if N is greater than 8, The pixels obtained for each NXN lattice are computed from a
  • the media processing device receives the image in units of 8 ⁇ 8 dots.
  • Image so, the obtaining unit 1001 obtains an 8 ⁇ 8 dot matrix that constitutes an image.
  • the media processing device can cache the 8X8 dot matrix. Therefore, the media processing device can also include a buffering unit with a buffer function (not shown in FIG. 10).
  • the obtaining unit 1001 can The 8X8 dot matrix is sent to the cache unit for caching.
  • the selecting unit 1002 can select the pixel points participating in the operation from the 8 ⁇ 8 dot matrix buffered by the buffer unit.
  • the media processing device described above may further include a storage unit 1004 for storing intermediate results obtained by the arithmetic unit 1003. After obtaining all the intermediate results constituting the NXN lattice, the arithmetic unit 1003 can perform all the intermediate results to obtain the pixel points corresponding to the NXN lattice.
  • the operation unit 1003 is specifically configured to perform an average operation on the pixels selected by the selection unit 1002.
  • the operation unit 1003 may be specifically configured to perform a summation operation on the pixel points selected by the selection unit 1002 from the 8 ⁇ 8 dot matrix currently obtained by the obtaining unit 1001.
  • the intermediate result corresponding to the NXN lattice is the result of the summation operation performed by the operation unit 1003 on the pixel points selected by the selection unit 1002 from the 8X8 dot matrix currently obtained by the obtaining unit 1001.
  • the operation unit 1003 may be specifically configured to perform a mean operation on the results of all the summation operations corresponding to the NXN lattice to obtain pixel points corresponding to the NXN lattice, wherein the denominator used in the mean operation is all the pixels participating in the operation. number.
  • the function and the mutual cooperation relationship of each unit in the foregoing media processing device can be referred to the method embodiment shown in FIG.
  • the present invention also provides an image processing system including a decoder, a display device, and the above-described media processing device.
  • the function of the media processing device and the cooperation relationship between the media processing device and the decoder and the display device can be referred to the above description, and details are not described herein again.
  • the media processing device selects from the currently obtained 8 ⁇ 8 dot matrix to participate in reducing the image signal in units of 8 ⁇ 8 dot matrix by N times.
  • the processed pixel points perform operations on the selected pixel points, output pixel points or intermediate results corresponding to the NXN dot matrix, and do not need to cache multiple 8 ⁇ 8 dot matrixes, and then perform reduction processing, thereby saving a plurality of 8 ⁇ 8 dot matrixes for buffering.
  • the required cache resources the media processing device performs calculation on the pixel points selected from the currently obtained 8 ⁇ 8 dot matrix to obtain pixel points or intermediate results, and does not need anti-aliasing filtering, and the implementation is relatively simple.
  • the solution can be embodied in the form of a software product that can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including a number of instructions for making a computer device (
  • a non-volatile storage medium which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.
  • the method described in various embodiments of the present invention may be a personal computer, a server, or a network device.

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Image Processing (AREA)

Abstract

The present invention provides an image downscaling method, which includes the following steps: a media processing device obtains an image signal in the units of 8×8 dot matrixes; the media processing device selects pixel points from the 8×8 dot matrix obtained presently, and the selected pixel points are used for the processing for downscaling the image signal in the units of 8×8 dot matrixes by a factor of N, wherein, N is a positive integer; the media processing device performs calculation on the pixel points selected from the 8×8 dot matrix obtained presently; if N<8, the pixel point corresponding to the N×N dot matrix is obtained; if N>8, the intermediate result corresponding to the N×N dot matrix is obtained; wherein, for each N×N dot matrix, one pixel point is obtained, and if N>8, the obtained pixel point for each N×N dot matrix is obtained by performing calculation on the multiple intermediate results. The invention also provides a media processing device and an image processing system.

Description

图像缩小方法和装置 技术领域  Image reduction method and device
本发明涉及图像处理技术， 尤其涉及图像缩小方法和装置。  The present invention relates to image processing techniques, and more particularly to an image reduction method and apparatus.
背景技术 Background technique
图像缩放技术被普遍的应用于各种媒体处理设备中。当图像信号进入媒体 处理设备后，媒体处理设备要将图像信号提供给显示设备。 由于显示设备的分 辨率往往与信号源的分辨率不同， 所以，媒体处理设备往往需要使用图像缩放 技术对图像信号进行缩放处理后，再将缩放后的图像信号提供给显示设备， 其 中， 本发明所述图像包括视频图像和图片。  Image scaling techniques are commonly used in a variety of media processing devices. When the image signal enters the media processing device, the media processing device provides the image signal to the display device. Since the resolution of the display device is often different from the resolution of the signal source, the media processing device often needs to use an image scaling technique to scale the image signal, and then provide the scaled image signal to the display device. The image includes a video image and a picture.
例如， 现在比较常见的高清图像分辨率是 1440X1080, 如果需要将分辨 率为 1440X1080的高清图像显示在分辨率为 640X480的显示设备，则需要将 分辨率为 1440X1080的图像缩小 2.25 (水平方向） X2.25 (竖直方向）倍后， 才能在分辨率为 640X480的显示设备中显示。  For example, the more common high-definition image resolution is 1440X1080. If you need to display a high-definition image with a resolution of 1440X1080 on a display device with a resolution of 640X480, you need to reduce the image with a resolution of 1440X1080 by 2.25 (horizontal direction) X2. After 25 (vertical) times, it can be displayed on a display device with a resolution of 640X480.
目前， 由于使用显示设备的用户的需求等原因， 图像有时需要以预览模式 显示在显示设备中。预览模式也要求媒体处理设备将获得的图像信号缩小到一 定比例后， 再将缩小后的图像信号提供给显示设备。 在实际应用中， 解码器将 被压缩的图像信号进行离散余弦变换（ DCT, Discrete Cosine Transform)解码， 被压缩的图像信号例如为运动图像专家组 （MPEG, Moving Picture Experts Group )格式的图像信号、 联合图像专家组（ JPEG, Joint Picture Expert Group ) 格式的图像信号等， 解码器将经过 DCT解码技术解码后的图像信号输入至媒 体处理设备，媒体处理设备对经过 DCT解码技术解码后的图像信号进行处理， 并将处理后的图像信号输入至显示设备显示。  Currently, images sometimes need to be displayed in a preview mode in a display device due to the needs of a user who uses a display device or the like. The preview mode also requires the media processing device to reduce the obtained image signal to a certain ratio, and then provide the reduced image signal to the display device. In a practical application, the decoder performs a discrete cosine transform (DCT, Discrete Cosine Transform) decoding on the compressed image signal, and the compressed image signal is, for example, an image signal in the format of the Moving Picture Experts Group (MPEG). The image signal of the Joint Picture Expert Group (JPEG) format, the decoder inputs the image signal decoded by the DCT decoding technology to the media processing device, and the media processing device performs the image signal decoded by the DCT decoding technology. Processing, and inputting the processed image signal to the display device display.
解码器一般要将经过 DCT解码技术解码后的图像信号以 8 X 8点阵为单位 输出， 媒体处理设备获得以 8X8点阵为单位的图像信号后， 一般要緩存多个 8X8点阵， 之后， 再以行为单位进行缩小处理。  The decoder generally outputs the image signal decoded by the DCT decoding technology in units of 8×8 dot matrix. After the media processing device obtains the image signal in units of 8×8 dot matrix, it is generally necessary to buffer a plurality of 8×8 dot matrix, and then, Then reduce the processing in units of actions.
然而， 在上述现有技术中， 媒体处理设备获得以 8X8点阵为单位的图像 信号后， 对图像信号进行缩小处理前， 需要緩存多个 8X8点阵， 这样， 媒体 处理设备需要緩存资源来緩存多个 8X8点阵， 这样的緩存资源对于媒体处理 设备来说是不小的资源开销。 另外， 对于图像信号的缩小处理， 传统的方法大 都是釆用多相位插值，具体方式是对多个样点的乘加运算得到一个像素点的输 出， 而且还需要加上抗混叠滤波， 这样的缩小处理方式较为复杂。 However, in the above prior art, after the media processing device obtains the image signal in units of 8×8 dot matrix, before the image signal is reduced, a plurality of 8×8 dot matrix needs to be cached, so that the media processing device needs to cache resources to cache. Multiple 8X8 dot matrix, such cache resources for media processing The device is not a small resource overhead. In addition, for the reduction processing of image signals, the traditional method mostly uses multi-phase interpolation, which is to multiply and add multiple samples to obtain the output of one pixel, and also needs anti-aliasing filtering. The reduction processing method is more complicated.
发明内容 Summary of the invention
本发明提供图像缩小方法和装置， 用以在对以 8X8点阵为单位的图像信 号进行缩小处理时， 节省媒体处理设备的緩存资源。  The present invention provides an image reduction method and apparatus for saving cache resources of a media processing device when performing image reduction processing on an image signal in units of 8×8 dots.
本发明提供一种图像缩小方法， 适用于图像预览模式，应用于包括解码器 和媒体处理设备的图像处理***， 媒体处理设备对解码器输出的以 8X8点阵 为单位的图像信号进行缩小处理， 所述以 8X8点阵为单位的图像信号是解码 器经过 DCT解码后得到的图像信号， 所述方法包括： 媒体处理设备获得以 8 X8点阵为单位的图像信号； 媒体处理设备从当前获得的 8X8点阵中选取像 素点， 用于参与将以 8X8点阵为单位的图像信号缩小 N倍的处理， 其中， N 为正整数；媒体处理设备对从当前获得的 8X8点阵中选取的像素点进行运算； 如果 N小于 8, 则得到对应 NXN点阵的像素点； 如果 N大于 8, 则得到对应 NXN点阵的中间结果； 其中， 对于每个 NXN点阵都得到一个像素点， 如果 N大于 8,则对于每个 NXN点阵得到的像素点由多个中间结果进行运算得到。  The invention provides an image reduction method, which is suitable for an image preview mode, and is applied to an image processing system including a decoder and a media processing device, and the media processing device reduces the image signal output by the decoder in units of 8×8 dot matrix. The image signal in the 8×8 dot matrix is an image signal obtained by the decoder after DCT decoding, and the method includes: the media processing device obtains an image signal in units of 8×8 dot matrix; the media processing device is currently obtained A pixel is selected in the 8X8 dot matrix for participating in a process of reducing the image signal in units of 8×8 dot matrix by N times, wherein N is a positive integer; the pixel selected by the media processing device from the currently obtained 8×8 dot matrix Performing an operation; if N is less than 8, the pixel corresponding to the NXN lattice is obtained; if N is greater than 8, an intermediate result corresponding to the NXN lattice is obtained; wherein, for each NXN lattice, one pixel is obtained, if N is greater than 8. The pixel points obtained for each NXN lattice are obtained by computing multiple intermediate results.
本发明还提供一种媒体处理设备， 适用于图像预览模式，应用于包括解码 器和媒体处理设备的图像处理*** , 媒体处理设备对解码器输出的以 8 X 8点 阵为单位的图像信号进行缩小处理， 所述以 8X8点阵为单位的图像信号是解 码器经过 DCT解码后得到的图像信号， 所述媒体处理设备包括： 获得单元， 用于获得以 8X8点阵为单位的图像信号； 选取单元， 用于从当前获得的 8X8 点阵中选取像素点， 选取的像素点用于参与将以 8X8点阵为单位的图像信号 缩小 N倍的处理， 其中， N为正整数； 运算单元， 用于对从当前获得的 8X8 点阵中选取的像素点进行运算； 如果 N小于 8, 则得到对应 NXN点阵的像素 点； 如果 N大于 8, 则得到对应 NXN点阵的中间结果； 其中， 对于每个 NX N点阵都得到一个像素点， 如果 N大于 8, 则对于每个 NXN点阵得到的像素 点由多个中间结果进行运算得到。  The present invention also provides a media processing device suitable for an image preview mode, applied to an image processing system including a decoder and a media processing device, wherein the media processing device performs an image signal output by the decoder in an 8×8 dot matrix unit. a reduction processing, the image signal in the unit of 8×8 dot matrix is an image signal obtained by the decoder after DCT decoding, and the media processing device includes: an obtaining unit, configured to obtain an image signal in units of 8×8 dot matrix; a unit, configured to select a pixel from the currently obtained 8×8 dot matrix, and the selected pixel is used to participate in a process of reducing an image signal of 8×8 dot matrix by N times, wherein N is a positive integer; Calculating the pixel points selected from the currently obtained 8×8 dot matrix; if N is less than 8, the pixel points corresponding to the NXN lattice are obtained; if N is greater than 8, the intermediate result corresponding to the NXN lattice is obtained; Each NX N dot matrix gets one pixel. If N is greater than 8, the pixel points obtained for each NXN dot matrix are composed of multiple intermediate nodes. Calculates get.
在本发明中， 媒体处理设备获得以 8X8点阵为单位的图像信号后， 从当 前获得的 8X8点阵中选取用于参与将以 8X8点阵为单位的图像信号缩小 N 倍的处理的像素点， 对选取的像素点进行运算， 输出对应 NXN点阵的像素点 或中间结果， 无须緩存多个 8X8点阵后， 再进行缩小处理， 节省了为緩存多 个 8X8点阵而需要的緩存资源。 In the present invention, after the media processing device obtains the image signal in units of 8×8 dot matrix, the image signal for participating in the 8×8 dot matrix is selected from the currently obtained 8×8 dot matrix to be reduced by N. Double the processed pixel points, calculate the selected pixel points, output the pixel points or intermediate results corresponding to the NXN dot matrix, and do not need to cache multiple 8X8 dot matrixes, then perform reduction processing, saving multiple 8X8 dot matrix for buffering And the cache resources needed.
附图说明 DRAWINGS
为了更清楚地说明本发明实施例或现有技术中的技术方案，下面将对实施 例或现有技术描述中所需要使用的附图作简单地介绍，显而易见地， 下面描述 中的附图仅仅是本发明的一些实施例，对于本领域普通技术人员来讲， 在不付 出创造性劳动的前提下， 还可以根据这些附图获得其他的附图。  In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the embodiments or the description of the prior art will be briefly described below. Obviously, the drawings in the following description are only It is a certain embodiment of the present invention, and other drawings can be obtained from those skilled in the art without any creative work.
图 1是本发明应用的图像处理***的一种逻辑结构示意图；  1 is a schematic diagram showing a logical structure of an image processing system to which the present invention is applied;
图 2是一个 8X8点阵的示意图；  Figure 2 is a schematic diagram of an 8X8 dot matrix;
图 3是本发明的一种图像缩小方法的流程图；  3 is a flow chart of an image reduction method of the present invention;
图 4A是本发明一个实施例中的缩小倍数为 2的情况下的一个 8X8点阵 被划分情况的示意图；  Figure 4A is a diagram showing the division of an 8X8 dot matrix in the case where the reduction factor is 2 in one embodiment of the present invention;
图 4B是本发明一个实施例中的缩小倍数为 2 的情况下的基本模板示意 图；  4B is a schematic diagram of a basic template in the case where the reduction factor is 2 in one embodiment of the present invention;
图 5A是本发明一个实施例中的缩小倍数为 4的情况下的一个 8X8点阵 被划分情况的示意图；  Figure 5A is a diagram showing the division of an 8X8 dot matrix in the case where the reduction factor is 4 in one embodiment of the present invention;
图 5B是本发明一个实施例中的缩小倍数为 4 的情况下的基本模板示意 图；  Figure 5B is a schematic diagram of a basic template in the case where the reduction factor is 4 in one embodiment of the present invention;
图 6A是本发明一个实施例中的缩小倍数为 3的情况下的 3X3个 8X8点 阵被划分情况的示意图；  Figure 6A is a diagram showing the division of 3X3 8X8 dot matrix in the case where the reduction factor is 3 in one embodiment of the present invention;
图 6B和图 6C是本发明一个实施例的缩小倍数为 3的情况下的基本模板 示意图；  6B and 6C are schematic diagrams showing a basic template in the case where the reduction factor is 3 in one embodiment of the present invention;
图 7是本发明一个实施例中的缩小倍数为 16的情况下的 16X16点阵的组 成示意图；  Figure 7 is a diagram showing the composition of a 16X16 dot matrix in the case where the reduction factor is 16 in one embodiment of the present invention;
图 8是本发明一个实施例中的缩小倍数为 12的情况下的多个 8X8点阵组 成多个 12X12点阵的示意图；  8 is a schematic diagram of a plurality of 8×8 dot matrix groups formed into a plurality of 12×12 dot matrix in the case where the reduction factor is 12 in one embodiment of the present invention;
图 9是本发明的图像缩小方法的一种实现流程图；  9 is a flow chart showing an implementation of the image reduction method of the present invention;
图 10是本发明的一种媒体处理设备的逻辑结构示意图。 具体实施方式 FIG. 10 is a schematic diagram showing the logical structure of a media processing device of the present invention. detailed description
下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行清 楚、 完整地描述， 显然， 所描述的实施例是本发明一部分实施例， 而不是全部 的实施例。基于本发明中的实施例， 本领域普通技术人员在没有作出创造性劳 动前提下所获得的所有其他实施例， 都属于本发明保护的范围。  BRIEF DESCRIPTION OF THE DRAWINGS The technical solutions in the embodiments of the present invention are clearly and completely described in the following with reference to the accompanying drawings. All other embodiments obtained by a person of ordinary skill in the art based on the embodiments of the present invention without creative efforts are within the scope of the present invention.
首先介绍本发明应用的图像处理***。  First, an image processing system to which the present invention is applied will be described.
如图 1所示， 图像处理***包括解码器、媒体处理设备和显示设备。 解码 器输出以 8X8点阵为单位的图像信号，媒体处理设备对以 8X8点阵为单位的 图像信号进行缩小处理， 并输出像素点给显示设备。  As shown in FIG. 1, the image processing system includes a decoder, a media processing device, and a display device. The decoder outputs an image signal in units of 8×8 dot matrix, and the media processing device performs a reduction process on the image signal in units of 8×8 dot matrix, and outputs the pixel point to the display device.
下面介绍本发明涉及的一些专业知识。  Some of the expertise involved in the present invention is described below.
如图 2所示， 一个 8 X 8点阵包括 64个像素点， 呈 8行、 8列分布， 每行 包括 8个像素点， 每列包括 8个像素点， 任意一个像素点在 8X8点阵中的位 置由这个像素点在 8X8点阵中所在的行和列来决定。假设 8X8点阵中的行由 i (i为不大于 8的正整数）来表示， 列由 j (j为不大于 8的正整数） 来表示， 那么 8X8点阵中的任意一个像素点可以用 A (i, j)来表示。  As shown in Figure 2, an 8 X 8 dot matrix consists of 64 pixels, distributed in 8 rows and 8 columns, each row consisting of 8 pixels, each column consisting of 8 pixels, and any pixel is in 8X8 dot matrix. The position in this is determined by the row and column of this pixel in the 8X8 dot matrix. Suppose the rows in the 8X8 dot matrix are represented by i (i is a positive integer not greater than 8), and the column is represented by j (j is a positive integer not greater than 8), then any pixel in the 8X8 dot matrix can be used. A (i, j) to indicate.
在本发明中， 媒体处理设备在接收以 8X8点阵为单位的图像信号时， 以 8 X 8点阵为单位进行接收， 即， 接收完第 1个 8 X 8点阵后， 再接收第 2个 8 X8点阵， 直到接收完第一个 8行的所有像素点后， 再接收第二个 8行的第 1 个 8x8点阵、 第 2个 8x8点阵， 直至接受完第二个 8行的所有像素点后， 再 接收第三个 8行的第 1个 8 X 8点阵、 第 2个 8 X 8点阵， 以此类推， 直至接收 完所有的 8X8点阵。  In the present invention, when receiving the image signal in units of 8×8 dot matrix, the media processing device receives the signal in units of 8×8 dots, that is, after receiving the first 8×8 dot matrix, and then receives the second image. 8 X8 dot matrix, after receiving all the pixels of the first 8 rows, then receive the first 8x8 dot matrix of the second 8 rows, the 2nd 8x8 dot matrix, until the second 8 rows are accepted After all the pixels, the first 8 X 8 dot matrix, the 2nd 8 X 8 dot matrix of the third 8 rows, and so on, until all 8X8 dot matrixes are received.
在本发明中， 对以 8X8点阵为单位的图像信号缩小 N倍的处理是指在水 平方向和竖直方向分别对图像信号缩小 N倍， 即， 对以 8X8点阵为单位的图 像信号缩小 N (水平方向） XN (竖直方向）倍， 对以 8X8点阵为单位的图 像信号缩小 N倍只是一种便于描述的方式而已。 对图像信号缩小 N倍可以认 为是对图像缩小 N倍， 也可以认为是对组成图像的每个 8X8点阵缩小 N倍， 其中， N是正整数。在本发明中，对于每个 NXN点阵都对应输出一个像素点。  In the present invention, the process of reducing the image signal by the unit of 8×8 dot matrix by N times is to reduce the image signal by N times in the horizontal direction and the vertical direction, that is, to reduce the image signal in units of 8×8 dot matrix. N (horizontal direction) XN (vertical direction) times, N-folding the image signal in units of 8×8 dot matrix is just a convenient way to describe it. N-folding the image signal can be considered as N-folding the image, or it can be considered to be N times smaller for each 8X8 dot matrix that makes up the image, where N is a positive integer. In the present invention, one pixel point is output correspondingly for each NXN dot matrix.
本发明下面的实施例中提到的基本模板是媒体处理设备用于选取像素点 的模板， 模板中规定从点阵中选取的像素点。 模板的设计方式不是唯一的， 或 者说，从点阵中选取哪些像素点不是唯一的， 具体可以由本领域技术人员根据 实际需要而设计。 The basic template mentioned in the following embodiments of the present invention is a template used by the media processing device to select pixels, and the template defines pixel points selected from the dot matrix. The design of the template is not unique, or It is said that which pixel points are selected from the dot matrix is not unique, and can be designed by a person skilled in the art according to actual needs.
另外， 对于彩色图像， 一个像素点的信息可以由多个分量表示， 例如 R、 In addition, for a color image, the information of one pixel can be represented by multiple components, such as R,
G、 8或丫、 Cb、 Cr等。 在对选取的像素点进行运算时， 应分别就该像素点每 个分量进行运算。 假设一个像素点对应 3个分量， 分别为分量 1、 分量 2和分 量 3 , 在对选取的像素点进行运算时， 应对该像素点的分量 1运算得到输出像 素点的分量 1 , 同理， 对该像素点的分量 2运算得到输出的像素点的分量 2, 对该像素点的分量 3运算得到输出的像素点的分量 3 , 这样就得到了输出的像 素点的 3个分量， 也就得到了输出的像素点。 G, 8 or 丫, Cb, Cr, etc. When performing operations on selected pixels, each component of the pixel should be operated separately. Suppose one pixel corresponds to three components, namely component 1, component 2 and component 3. When performing the operation on the selected pixel, the component 1 of the pixel should be calculated to obtain the component 1 of the output pixel. Similarly, The component 2 of the pixel is calculated to obtain the component 2 of the output pixel, and the component 3 of the pixel is calculated to obtain the component 3 of the output pixel, so that three components of the output pixel are obtained, and thus The pixel point of the output.
本发明所有的实施例都适用于图像预览模式。本发明所有的实施例都可以 应用于包括解码器和媒体处理设备的图像处理***，媒体处理设备对解码器输 出的以 8 X 8点阵为单位的图像信号进行缩小处理，以 8 X 8点阵为单位的图像 信号是解码器经过 DCT解码后得到的图像信号。 下面介绍本发明的一种图像缩小方法，这种方法可以从媒体处理设备的角 度来描述。 如图 3所示， 这种方法包括：  All of the embodiments of the present invention are applicable to the image preview mode. All embodiments of the present invention can be applied to an image processing system including a decoder and a media processing device, and the media processing device performs a reduction process on the image signal of the 8×8 dot matrix output by the decoder to 8×8 points. The image signal of the array is the image signal obtained by the decoder after DCT decoding. An image reduction method of the present invention will now be described, which can be described from the perspective of a media processing device. As shown in Figure 3, this method includes:
S301 : 媒体处理设备获得以 8 X 8点阵为单位的图像信号。  S301: The media processing device obtains an image signal in units of 8×8 dots.
S302: 媒体处理设备从当前获得的 8 X 8点阵中选取像素点， 用于参与将 以 8 X 8点阵为单位的图像信号缩小 N倍的处理， 其中， N为正整数。  S302: The media processing device selects a pixel from the currently obtained 8×8 dot matrix for participating in a process of reducing an image signal of 8×8 dot matrix by N times, wherein N is a positive integer.
S303:媒体处理设备对从当前获得的 8 X 8点阵中选取的像素点进行运算； 如果 N小于 8, 则得到对应 NXN点阵的像素点； 如果 N大于 8, 则得到对应 NXN点阵的中间结果； 其中， 对于每个 NXN点阵都得到一个像素点， 如果 N大于 8,则对于每个 NXN点阵得到的像素点由多个中间结果进行运算得到。  S303: The media processing device performs operations on the pixels selected from the currently obtained 8×8 dot matrix; if N is less than 8, the pixel points corresponding to the NXN lattice are obtained; if N is greater than 8, the corresponding NXN lattice is obtained. Intermediate result; wherein, for each NXN lattice, one pixel is obtained. If N is greater than 8, the pixel obtained for each NXN lattice is calculated by using multiple intermediate results.
媒体处理设备对从当前获得的 8 X 8点阵中选取的像素点进行运算时， 可 以对选取的像素点进行均值运算。  When the media processing device operates on the pixels selected from the currently obtained 8 X 8 dot matrix, the selected pixel points can be averaged.
如果媒体处理设备得到对应 NXN点阵的中间结果，那么媒体处理设备可 以存储所述中间结果。 媒体处理设备在得到对应 NXN 点阵的所有中间结果 后， 对所有的中间结果进行运算， 可以得到对应 NXN点阵的像素点。  If the media processing device gets an intermediate result corresponding to the NXN lattice, the media processing device can store the intermediate result. After obtaining all the intermediate results of the corresponding NXN lattice, the media processing device calculates all the intermediate results and obtains the pixels corresponding to the NXN lattice.
媒体处理设备对从当前获得的 8 X 8点阵中选取的像素点进行运算时， 可 以对从当前获得的 8X8点阵中选取的像素点进行求和运算。 这种情况下， 所 述对应 NXN点阵的中间结果即为媒体处理设备对从当前获得的 8X8点阵中 选取的像素点进行求和运算的结果。媒体处理设备对所有的中间结果进行运算 时， 可以对所有的求和运算的结果进行均值运算，从而得到对应 NXN点阵的 像素点。具体的，如果 N是 8的约数，那么一个 8X8点阵可以被划分为（8/N) (水平方向） X (8/N) (竖直方向）个 NXN点阵。 When the media processing device operates on the pixels selected from the currently obtained 8×8 dot matrix, The summation is performed on the pixels selected from the currently obtained 8×8 dot matrix. In this case, the intermediate result of the corresponding NXN lattice is the result of the summation operation performed by the media processing device on the pixel points selected from the currently obtained 8×8 dot matrix. When the media processing device operates on all the intermediate results, the result of all the summation operations can be averaged to obtain the pixel points corresponding to the NXN lattice. Specifically, if N is a divisor of 8, then an 8×8 dot matrix can be divided into (8/N) (horizontal direction) X (8/N) (vertical direction) NXN lattices.
这种情况下， 媒体处理设备可以分别从当前获得的 8X8点阵中的每个 N X N点阵中选取像素点， 并分别对从当前获得的 8 X 8点阵中的每个 N X N点 阵中选取的像素点进行运算， 得到对应当前获得的 8X8点阵中的每个 NXN 点阵的像素点。  In this case, the media processing device may respectively select pixels from each of the currently obtained 8×8 lattices, and respectively select each of the NXN lattices from the currently obtained 8×8 dot matrix. The pixels are operated to obtain pixels corresponding to each of the currently obtained 8×8 lattices.
以 N=2为例， 媒体处理设备需要将接收到的图像信号缩小 2倍。 如图 4A 所示， 一个 8X8点阵可以被划分为 4X4个 2X2点阵，每个 2X2点阵都输出 一个像素点， 输出的像素点记为 0 (i, j)。 在这个实施例中， 基本模板如图 4B所示， 由图 4B可知， 每个 2X2点阵中的所有像素点都参与输出像素点的 运算， 运算方式为取基本模板中所有的像素点的值的平均值， 即从 2X2点阵 中选取所有的像素点做均值运算。 由此可以依次得到一个 8X8点阵输出的 4 X 4个像素点， 分别如下：  Taking N=2 as an example, the media processing device needs to reduce the received image signal by a factor of two. As shown in Fig. 4A, an 8X8 dot matrix can be divided into 4X4 2X2 dot matrixes, each of which outputs one pixel point, and the output pixel points are recorded as 0 (i, j). In this embodiment, the basic template is as shown in FIG. 4B. As can be seen from FIG. 4B, all the pixels in each 2×2 dot matrix participate in the operation of outputting pixel points, and the operation manner is to take the values of all the pixel points in the basic template. The average value, that is, all the pixels are selected from the 2X2 dot matrix for the mean operation. Thus, 4 X 4 pixel points of an 8X8 dot matrix output can be obtained in sequence, as follows:
0(1， 1) = [A(l， 1) + A(1， 2) +A(2, 1)+A(2， 2)]/ 4  0(1, 1) = [A(l, 1) + A(1, 2) +A(2, 1)+A(2, 2)]/ 4
0(1， 2) = [A(l， 3) + A(l， 4)+A(2， 3)+A(2， 4) ]/4 0(1, 2) = [A(l, 3) + A(l, 4)+A(2, 3)+A(2, 4) ]/4
0(4， 3) = [A(7， 5) + A(7， 6)+A(8， 5)+A(8， 6)]/ 4 0(4, 3) = [A(7, 5) + A(7, 6)+A(8, 5)+A(8, 6)]/ 4
0(4， 4) = [A(7， 7) + A(7， 8)+A(8， 7)+A(8， 8)]/ 4 0(4, 4) = [A(7, 7) + A(7, 8)+A(8, 7)+A(8, 8)]/ 4
再以 N=4为例， 媒体处理设备需要将接收到的图像信号缩小 4倍。 如图 5A所示， 一个 8X8点阵可以被划分为 2X2个 4X4点阵， 每个 4X4点阵都 输出一个像素点， 输出的像素点记为 0 (i, j)。 在这个实施例中， 基本模板如 图 5B所示， 每个 4X4点阵中的阴影像素点参与输出像素点的运算， 运算方 式为取基本模板中阴影像素点的值的平均值。 由此可以依次得到一个 8X8点 阵输出的 2X2个像素点， 分别如下： 0(1， 1) = [A(l, 1) + A(1, 3)+A(3, 1)+A(3, 3)]/ 4 Taking N=4 as an example, the media processing device needs to reduce the received image signal by 4 times. As shown in Fig. 5A, an 8×8 dot matrix can be divided into 2×2 4×4 dot matrixes, each of which outputs one pixel point, and the output pixel points are recorded as 0 (i, j). In this embodiment, the basic template is as shown in FIG. 5B, and the shadow pixel points in each 4×4 dot matrix participate in the operation of outputting pixel points in such a manner as to take the average value of the values of the shaded pixel points in the basic template. Thus, 2X2 pixels of an 8X8 dot matrix output can be obtained in sequence, as follows: 0(1, 1) = [A(l, 1) + A(1, 3)+A(3, 1)+A(3, 3)]/ 4
0(1， 2) = [A(l， 5) + A(l， 7) +A (3， 5)+A(3， 7)]/ 4 0(2， 1) = [A(5， 1) + A(5， 3)+A(7， 1)+A(7， 3)]/ 4 0(1, 2) = [A(l, 5) + A(l, 7) +A (3, 5)+A(3, 7)]/ 4 0(2, 1) = [A(5, 1) + A(5, 3)+A(7, 1)+A(7, 3)]/ 4
0(2， 2) = [A(5， 5)+A(5， 7) +A(7, 5)+A(7， 7)]/ 4 0(2, 2) = [A(5, 5)+A(5, 7) +A(7, 5)+A(7, 7)]/ 4
如果 N小于 8且不是 8的约数， 那么一个 8X8点阵可以被划分为至少一 个 NXN点阵和多个边界像素点， 一个 8X8点阵中的边界像素点可以与相邻 的 8X8点阵中的边界像素点组成 NXN点阵， 其中， 这里的边界像素点是指 8X8点阵中的除 NXN点阵包括的像素点之外的像素点， 或者说， 8X8点阵 中与相邻 8 X 8点阵中的像素点一同参与组成 N X N点阵的像素点。  If N is less than 8 and not a divisor of 8, then an 8X8 lattice can be divided into at least one NXN lattice and a plurality of boundary pixels, and boundary pixels in an 8×8 lattice can be associated with adjacent 8×8 lattices. The boundary pixels constitute an NXN lattice, wherein the boundary pixel here refers to a pixel in the 8×8 dot matrix other than the pixel included in the NXN dot matrix, or 8×8 dot matrix and adjacent 8 X 8 The pixels in the dot matrix participate in the pixel points that make up the NXN dot matrix.
这种情况下， 媒体处理设备可以分别从当前获得的 8X8点阵中的每个 N In this case, the media processing device can each from each of the currently acquired 8X8 lattices.
X N点阵中选取像素点， 并分别对从当前获得的 8 X 8点阵中的每个 N X N点 阵中选取的像素点进行运算， 分别得到对应当前获得的 8X8点阵中的每个 N XN点阵的像素点。 如果当前获得的 8X8点阵中用于组成一个 NXN点阵的 边界像素点的数量比相邻的 8X8点阵中用于组成所述 NXN点阵的边界像素 点的数量多， 那么媒体处理设备可以从当前获得的 8X8点阵中用于组成所述 NXN点阵的边界像素点中选取像素点， 并对从当前获得的 8X8点阵中用于 组成所述 NXN点阵的边界像素点中选取的像素点进行运算，得到对应所述 N XN点阵的像素点。 Selecting pixels in the XN lattice, and respectively computing the pixels selected from each of the currently obtained 8×8 lattices, respectively obtaining respective N XNs corresponding to the currently obtained 8×8 lattices. The pixel points of the dot matrix. If the number of boundary pixels used to form an NXN lattice in the currently obtained 8×8 dot matrix is larger than the number of boundary pixel dots used to form the NXN lattice in the adjacent 8×8 dot matrix, then the media processing device can Selecting pixel points from boundary pixels used to form the NXN lattice in the currently obtained 8×8 dot matrix, and selecting from the currently selected 8×8 dot matrix used to form the boundary pixels of the NXN lattice The pixel is operated to obtain a pixel corresponding to the N XN dot matrix.
以 N=3为例， 媒体处理设备需要将接收到的图像信号缩小 3倍。 这种情 况下， 一个 8 X 8点阵需要输出 2.67 X 2.67个像素点， 但在实际应用中， 没有 非整数个数的像素点。 对此， 取 3和 8的最小公倍数为 24, 这样， 3X3个 8 X8点阵输出 8X8个像素点。 如图 6A所示， 3X3个 8X8点阵可以被划分为 8X8个 3X3点阵， 每个 3X3点阵都输出一个像素点， 输出的像素点记为 0 a」)。  Taking N=3 as an example, the media processing device needs to reduce the received image signal by a factor of three. In this case, an 8 X 8 dot matrix needs to output 2.67 X 2.67 pixels, but in practice, there are no non-integer number of pixels. For this, the least common multiple of 3 and 8 is 24, so that 3X3 8 X8 dot matrix outputs 8X8 pixels. As shown in Fig. 6A, 3X3 8X8 dots can be divided into 8X8 3X3 dot matrix, each 3X3 dot matrix outputs one pixel, and the output pixel is recorded as 0 a").
这里需要说明的是， 第 1个 8X8点阵中的像素点、第 2个 8X8点阵中的 像素点和第 3个 8X8点阵中的像素点属于同一个 8行的像素点， 同理， 第 4 个 8X8点阵中的像素点、 第 5个 8X8点阵中的像素点和第 6个 8X8点阵中 的像素点属于同一个 8行的像素点， 第 7个 8X8点阵中的像素点、 第 8个 8 X 8点阵中的像素点和第 9个 8 X 8点阵中的像素点属于同一个 8行的像素点。 第 1个 8 X 8点阵的相邻 8 X 8点阵分别是第 2个 8 X 8点阵和第 4个 8 X 8点阵， 第 5个 8X8点阵的相邻 8X8点阵分别是第 1个 8X8点阵、第 2个 8X8点阵、 第 3个 8X8点阵、 第 4个 8X8点阵、 第 6个 8X8点阵、 第 7个 8X8点阵、 第 8个 8X8点阵和第 9个 8X8点阵，其他 8X8点阵的相邻 8X8点阵可以依 此确定， 这里不再赘述。 It should be noted that the pixels in the first 8×8 dot matrix, the pixels in the second 8×8 dot matrix, and the pixels in the third 8×8 dot matrix belong to the same 8-row pixel point. Similarly, The pixel in the 4th 8X8 dot matrix, the pixel in the 5th 8X8 dot matrix, and the pixel in the 6th 8X8 dot matrix belong to the same 8 rows of pixels, and the pixels in the 7th 8X8 dot matrix Point, 8th 8 The pixel points in the X 8 dot matrix and the pixel dots in the 9th 8 X 8 dot matrix belong to the same 8 rows of pixel points. The adjacent 8 X 8 lattices of the first 8 X 8 dot matrix are the 2nd 8 X 8 dot matrix and the 4th 8 X 8 dot matrix respectively, and the 5X 8 8 dot matrix adjacent 8X8 dot matrix are respectively 1st 8X8 dot matrix, 2nd 8X8 dot matrix, 3rd 8X8 dot matrix, 4th 8X8 dot matrix, 6th 8X8 dot matrix, 7th 8X8 dot matrix, 8th 8X8 dot matrix and 9 8X8 dot matrixes, and other 8X8 dot matrix adjacent 8X8 dot matrix can be determined accordingly, and will not be described here.
还需要说明的是， 如果媒体处理设备接收到的图像的像素点超过 24列， 那么媒体处理设备在分别接受完第 1个 8X8点阵、 第 2个 8X8点阵和第 3 个 8X8点阵后， 还需要依次接收同一行的其他 8X8点阵， 之后， 才会依次接 收第 4个 8X8点阵、 第 5个 8X8点阵和第 6个 8X8点阵， 同理， 在分别接 受完第 4个 8X8点阵、 第 5个 8X8点阵和第 6个 8X8点阵后， 还需要依次 接收与这 3个 8X8点阵同一行的其他 8X8点阵， 之后， 才会依次接收第 7 个 8X8点阵、 第 8个 8X8点阵和第 9个 8X8点阵。  It should also be noted that if the image received by the media processing device has more than 24 pixels, the media processing device after receiving the first 8×8 dot matrix, the second 8×8 dot matrix, and the third 8×8 dot matrix respectively. It is also necessary to receive the other 8X8 lattices in the same row in turn, and then receive the 4th 8X8 dot matrix, the 5th 8X8 dot matrix, and the 6th 8X8 dot matrix in turn, in the same way, after accepting the 4th respectively. After 8X8 dot matrix, 5th 8X8 dot matrix and 6th 8X8 dot matrix, you need to receive the other 8X8 dot matrix in the same row as the 3 8X8 dot matrix in turn, and then receive the 7th 8X8 dot matrix in turn. , 8th 8X8 dot matrix and 9th 8X8 dot matrix.
在这个实施例中，第 1个 8X8点阵的基本模板分别如图 6B和图 6C所示， 第 1个 8X8点阵输出 3X3个像素点， 3X3个像素点依次为：  In this embodiment, the basic templates of the first 8×8 dot matrix are respectively shown in FIG. 6B and FIG. 6C, and the first 8×8 dot matrix outputs 3×3 pixel points, and the 3×3 pixel points are:
0(1， 1) = [A(2， 1)+A(2， 3)]/2  0(1, 1) = [A(2, 1)+A(2, 3)]/2
0(1， 2) = [A(2， 4)+A(2， 6)]/2  0(1, 2) = [A(2, 4)+A(2, 6)]/2
0(1， 3) = [A(2， 7)+A(2, 8)]/2  0(1, 3) = [A(2, 7)+A(2, 8)]/2
0(2， 1) = [A(5， 1)+A(5， 3)]/2  0(2, 1) = [A(5, 1)+A(5, 3)]/2
0(2， 2) = [A(5， 4)+A(5， 6)]/2  0(2, 2) = [A(5, 4)+A(5, 6)]/2
0(2， 3) = [A(5， 7)+A(5， 8)]/2  0(2, 3) = [A(5, 7)+A(5, 8)]/2
0(3， 1) = [A(8， 1)+A(8， 3)]/2  0(3, 1) = [A(8, 1)+A(8, 3)]/2
0(3， 2) = [A(8， 4)+A(8， 6)]/2  0(3, 2) = [A(8, 4)+A(8, 6)]/2
0(3， 3) = [A(8， 7)+A(8， 8)]/2  0(3, 3) = [A(8, 7)+A(8, 8)]/2
需要说明的是， 第 1个 8X8点阵的第 7、 8列中的像素点与第 2个 8X8 点阵的第 1列中的像素点可以组合为 3个 3X3点阵，由于第 1个 8X8点阵的 第 7、 8列中的像素点数量较多， 所以， 从第 1个 8X8点阵的第 7、 8列中的 像素点中选择像素点参与输出像素点的运算。 同理， 第 1个 8X8点阵的第 7、 8行中的像素点与第 4个 8X8点阵的第 1行中的像素点可以组合为 3个 3X3 点阵， 由于第 1个 8X8点阵的第 7、 8行中的像素点数量较多， 所以， 从第 1 个 8X8点阵的第 7、 8行中的像素点中选择像素点参与输出像素点的运算。 It should be noted that the pixel points in the seventh and eighth columns of the first 8×8 dot matrix and the pixel points in the first column of the second 8×8 dot matrix can be combined into three 3×3 dot matrixes, due to the first 8×8 The number of pixels in the 7th and 8th columns of the dot matrix is large, so from the 7th and 8th columns of the 1st 8X8 dot matrix The pixel is selected among the pixels to participate in the operation of outputting the pixel. Similarly, the pixels in the 7th and 8th rows of the 1st 8X8 dot matrix and the pixels in the 1st row of the 4th 8X8 dot matrix can be combined into 3 3X3 dot matrixes, due to the first 8X8 dot matrix. In the seventh and eighth rows, the number of pixels is large. Therefore, the pixel point is selected from the pixels in the seventh and eighth rows of the first 8×8 dot matrix to participate in the operation of outputting the pixel point.
在这个实施例中， 第 2个 8X8点阵的基本模板如图 6B所示。 由于第 1 个 8X8点阵的第 7、 8列中的像素点已经输出 3个像素点， 所以， 第 2个 8X 8点阵的第 1列中的像素点在第 2个 8X8点阵中不参与输出像素点的运算。 同理， 第 2个 8X8点阵的第 8列中的像素点与第 3个 8X8点阵的第 1、 2列 中的像素点也可以组合为 3个 3X3点阵， 由于第 3个 8X8点阵的第 1、 2列 中的像素点数量较多， 所以， 第 3个 8X8点阵的第 1、 2列中的像素点输出 3 个像素点，第 2个 8X8点阵的第 8列中的像素点在第 2个 8X8点阵中不参与 输出像素点的运算。  In this embodiment, the basic template of the second 8X8 dot matrix is as shown in Fig. 6B. Since the pixels in the 7th and 8th columns of the 1st 8X8 dot matrix have already output 3 pixels, the pixel in the 1st column of the 2nd 8X 8 dot matrix is not in the 2nd 8X8 dot matrix. Participate in the operation of outputting pixel points. Similarly, the pixel in the eighth column of the second 8×8 dot matrix and the pixel in the first and second columns of the third 8×8 dot matrix can also be combined into three 3×3 dot matrix, due to the third 8×8 dot. The number of pixels in the first and second columns of the array is large. Therefore, the pixels in the first and second columns of the third 8×8 dot matrix output 3 pixels, and the eighth column of the second 8×8 dot matrix. The pixel points do not participate in the operation of outputting pixel points in the second 8×8 dot matrix.
因此， 第 2个 8X8点阵输出 2X3个像素点， 2X3个像素点分别如下： Therefore, the second 8X8 dot matrix output 2X3 pixels, 2X3 pixels are as follows:
0(1， 4) = [A(2， 2)+A(2, 4)]/2 0(1, 4) = [A(2, 2)+A(2, 4)]/2
0(1， 5) = [A(2， 5)+A(2， 7)]/2 0(1, 5) = [A(2, 5)+A(2, 7)]/2
0(2， 4) = [A(5， 2)+A(5， 4)]/2  0(2, 4) = [A(5, 2)+A(5, 4)]/2
0(2， 5) = [A(5， 5)+A(5， 7)]/2  0(2, 5) = [A(5, 5)+A(5, 7)]/2
0(3， 4) = [A(8， 2)+A(8， 4)]/2  0(3, 4) = [A(8, 2)+A(8, 4)]/2
0(1， 5) = [A(8， 5)+A(8， 7)]/2  0(1, 5) = [A(8, 5)+A(8, 7)]/2
在这个实施例中，第 4个 8 X 8点阵的基本模板分别如图 6B和图 6C所示。 由于第 1个 8X8点阵的第 7、 8行中的像素点已经输出 3个像素点， 所以， 第 4个 8 X 8点阵的第 1行中的像素点在该 8 X 8点阵中不参与输出像素点的运算。 同理， 该 8X8点阵的第 8行中的像素点与第 7个 8X8点阵的第 1、 2行中的 像素点也可以组合为 3个 3X3点阵， 由于第 7个 8X8点阵的第 7、 8行中的 像素点数量较多， 所以， 第 7个 8X8点阵的第 7、 8行中的像素点输出 3个像 素点，第 4个 8X8点阵的第 8行中的像素点在该 8X8点阵中不参与输出像素 点的运算。  In this embodiment, the basic templates of the fourth 8 X 8 dot matrix are shown in Figs. 6B and 6C, respectively. Since the pixels in the 7th and 8th rows of the 1st 8X8 dot matrix have already output 3 pixels, the pixel in the 1st row of the 4th 8×8 dot matrix is in the 8×8 dot matrix. Does not participate in the operation of outputting pixel points. Similarly, the pixels in the 8th row of the 8X8 dot matrix and the pixels in the 1st and 2nd rows of the 7th 8X8 dot matrix can also be combined into three 3X3 dot matrixes, due to the 7th 8X8 dot matrix In the 7th and 8th rows, the number of pixels is large. Therefore, the pixels in the 7th and 8th rows of the 7th 8X8 dot matrix output 3 pixels, and the pixels in the 8th row of the 4th 8X8 dot matrix. The point does not participate in the operation of outputting pixel points in the 8X8 dot matrix.
因此， 第 4个 8X8点阵输出 3X2个像素点， 3X2个像素点的值分别如 0(4， 1) = [A(3， 1)+A(3， 3)]/2 Therefore, the 4th 8X8 dot matrix outputs 3X2 pixels, and the values of 3X2 pixels are respectively 0(4, 1) = [A(3, 1)+A(3, 3)]/2
0(4， 2) = [A(3， 4)+A(3， 6)]/2 0(4, 2) = [A(3, 4)+A(3, 6)]/2
0(4， 3) = [A(3， 7)+A(3， 8)]/2  0(4, 3) = [A(3, 7)+A(3, 8)]/2
0(5， 1) = [A(6， 1)+A(6， 3)]/2  0(5, 1) = [A(6, 1)+A(6, 3)]/2
0(5， 2) = [A(6， 4)+A(6， 6)]/2  0(5, 2) = [A(6, 4)+A(6, 6)]/2
0(5， 3) = [A(6， 7)+A(6， 8)]/2  0(5, 3) = [A(6, 7)+A(6, 8)]/2
其他 8 X 8点阵输出的像素点的值与上述计算方法同理， 这里不再赘述。 再以 N=5为例， 媒体处理设备需要将接收到的图像信号缩小 5倍。 这种 情况下， 一个 8X8点阵需要输出 1.6X1.6个像素点， 但在实际应用中， 没有 非整数个数的像素点。 对此， 取 5和 8的最小公倍数为 40, 这样， 5X5个 8 X 8点阵输出 8 X 8个像素点。在这个实施例中， 5X5个 8X8点阵可以被划分 为 8X8个 5X5点阵，每个 5X5点阵都输出一个像素点。每个 5X5点阵输出 的像素点的值的运算可以参照 N=3时的处理情况， 这里不再赘述。  The values of the pixels of the other 8 X 8 dot matrix output are the same as those of the above calculation method, and will not be described here. Taking N=5 as an example, the media processing device needs to reduce the received image signal by 5 times. In this case, an 8X8 dot matrix needs to output 1.6X1.6 pixels, but in practical applications, there are no non-integer number of pixels. For this, the least common multiple of 5 and 8 is 40, so that 5X5 8 X 8 dot matrix outputs 8 X 8 pixels. In this embodiment, 5X5 8X8 dots can be divided into 8X8 5X5 dots, and each 5X5 dot matrix outputs one pixel. The operation of the value of the pixel of each 5X5 dot matrix output can refer to the processing situation when N=3, and will not be described here.
如果 N为 8的倍数，那么 kXk个 8X8点阵组成 NXN点阵，其中， k=N/8。 这种情况下， 媒体处理设备可以得到中间结果， 中间结果与当前获得的 8X8 点阵参与组成的 N X N点阵相对应。  If N is a multiple of 8, then kXk 8X8 lattices form the NXN lattice, where k = N/8. In this case, the media processing device can get an intermediate result, and the intermediate result corresponds to the N X N dot matrix composed of the currently acquired 8X8 dot matrix.
如果 N为 8的倍数， 那么媒体处理设备得到对应 NXN点阵的中间结果 之后， 可以存储中间结果。媒体处理设备在得到对应 NXN点阵的所有中间结 果后，可以对所有的中间结果进行运算，得到对应 NXN点阵的像素点。例如， 媒体处理设备可以对从当前获得的 8X8点阵中选取的像素点进行求和运算， 求和运算的结果即为中间结果，媒体处理设备可以对所有的求和运算的结果进 行均值运算， 得到对应 NXN点阵的像素点， 其中， 均值运算所使用的分母为 所有参与运算的像素点的个数， 即， 当前获得的 8X8点阵参与组成的 NXN 点阵中所有参与运算的像素点的个数。  If N is a multiple of 8, then the intermediate processing result can be stored after the media processing device gets an intermediate result corresponding to the NXN lattice. After obtaining all the intermediate results of the corresponding NXN lattice, the media processing device can calculate all the intermediate results to obtain the pixels corresponding to the NXN lattice. For example, the media processing device may perform a summation operation on the pixels selected from the currently obtained 8×8 dot matrix, and the result of the summation operation is an intermediate result, and the media processing device may perform an average operation on the results of all the summation operations. Obtaining a pixel corresponding to the NXN lattice, wherein the denominator used in the mean operation is the number of pixels participating in the operation, that is, the currently acquired 8×8 dot matrix participates in all the pixels of the NXN lattice participating in the operation. Number.
以 N=16为例， 媒体处理设备需要将接收到的图像信号缩小 16倍。 这种 情况下， 一个 8X8点阵需要输出 0.5X0.5个像素点， 但在实际应用中， 没有 非整数个数的像素点。 对此， 取 16和 8的最小公倍数为 16, 这样， 2X2个 8 X 8点阵输出 2X2个像素点， 具体划分如图 7所示。 在这个实施例中， 每个 2 X2个 8X8点阵都输出一个像素点。 本领域技术人员可以根据实际需要， 设 计出参与输出像素点的运算的像素点，使用平均值运算方式，得到输出的像素 点的值。 需要说明的是， 图 7中的 8X8点阵所属的行、 8X8点阵之间的相邻 关系以及接收 8 X 8点阵的顺序可以参见上述 N=3的实施例中的相关描述， 这 里不再赘述。 Taking N=16 as an example, the media processing device needs to reduce the received image signal by 16 times. In this case, an 8X8 dot matrix needs to output 0.5X0.5 pixels, but in practical applications, there are no non-integer number of pixels. For this, take the least common multiple of 16 and 8 to 16, so that 2X2 8 X 8 dot matrix output 2X2 pixels, the specific division is shown in Figure 7. In this embodiment, each 2 X2 8X8 dot matrix outputs one pixel. A person skilled in the art can design a pixel point participating in an operation of outputting a pixel point according to actual needs, and use an average value calculation method to obtain a value of the output pixel point. It should be noted that the order of the 8X8 dot matrix in FIG. 7 , the adjacent relationship between the 8×8 dot matrix, and the order of receiving the 8×8 dot matrix can be referred to the related description in the above N=3 embodiment, and Let me repeat.
需要说明的是，由于媒体处理设备是按照先后顺序分别接收到图 7所示的 4个 8X8点阵的， 所以， 如果 4个 8X8点阵中均有像素点需要参与输出像素 点的运算， 那么媒体处理设备在获得第 1个 8X8点阵后， 可以将第 1个 8X8 点阵中参与输出像素点的运算的像素点的值相加， 緩存结果， 在获得第 2个 8 X8点阵后， 可以将第 2个 8X8点阵中参与输出像素点的运算的像素点的值 相加， 緩存结果， 在获得第 3个 8X8点阵后， 可以将第 3个 8X8点阵中参与 输出像素点的运算的像素点的值相加，緩存结果， 在获得第 4个 8X8点阵后， 可以将第 4个 8X8点阵中参与输出像素点的运算的像素点的值相加， 得到一 个结果， 最后，将上述 4个结果做平均值运算，得到输出的像素点的值，其中， 平均值运算所使用的分母为第 1个 8X8点阵、 第 2个 8X8点阵、 第 3个 8X 8点阵和第 4个 8X8点阵中所有参与运算的像素点的个数。  It should be noted that, since the media processing device receives the four 8×8 dot matrixes shown in FIG. 7 in sequence, therefore, if all the pixels in the four 8×8 dot matrix need to participate in the operation of outputting pixel points, then After obtaining the first 8×8 dot matrix, the media processing device may add the values of the pixels participating in the operation of the output pixel in the first 8×8 dot matrix, and buffer the result. After obtaining the second 8×8 dot matrix, The values of the pixels participating in the operation of the output pixel in the second 8×8 dot matrix can be added, and the result can be cached. After the third 8×8 dot matrix is obtained, the third 8×8 dot matrix can participate in the output pixel point. The values of the calculated pixel points are added, and the result of the buffering is added. After obtaining the fourth 8×8 dot matrix, the values of the pixel points of the operation of the output pixel point in the fourth 8×8 dot matrix can be added to obtain a result, and finally The average of the above four results is obtained to obtain the value of the output pixel. The denominator used in the average calculation is the first 8×8 dot matrix, the second 8×8 dot matrix, and the third 8×8 dot matrix. And the 4th 8X8 The number of pixels in the lattice that participate in the operation.
如果 N大于 8且不是 8的倍数， 那么一个 NXN点阵由至少一个 8X8点 阵和相邻的 8X8 点阵中的边界像素点组成， 其中， 这里的像素点是指 8X8 点阵中与相邻的 8 X 8点阵中的像素点一同参与组成 N X N点阵的像素点。  If N is greater than 8 and not a multiple of 8, then an NXN lattice consists of at least one 8×8 dot matrix and boundary pixels in an adjacent 8×8 dot matrix, where the pixel points are adjacent to the 8×8 dot matrix. The pixels in the 8 X 8 dot matrix participate in the pixels that make up the NXN dot matrix.
如果当前获得的 8X8点阵中的所有像素点都参与组成 NXN点阵， 那么 媒体处理设备可以从当前获得的 8X8点阵中的所有像素点中选取像素点， 并 对从当前获得的 8X8点阵中的所有像素点中选取的像素点进行运算， 得到与 当前获得的 8 X 8点阵参与组成的 N X N点阵对应的中间结果。  If all the pixels in the currently acquired 8X8 dot matrix are involved in composing the NXN dot matrix, the media processing device can select pixels from all the pixels in the currently obtained 8×8 dot matrix, and the 8X8 dot matrix obtained from the current one. The pixel selected from all the pixels in the operation is operated to obtain an intermediate result corresponding to the NXN lattice composed of the currently obtained 8×8 lattice.
如果当前获得的 8X8点阵中的边界像素点参与组成 NXN点阵， 则媒体 处理设备从当前获得的 8X8点阵中的边界像素点中选取像素点， 并对从当前 获得的 8X8点阵中的边界像素点中选取的像素点进行运算， 得到与当前获得 的 8 X 8点阵参与组成的 N X N点阵对应的中间结果。  If the boundary pixel points in the currently obtained 8×8 dot matrix participate in the composition of the NXN dot matrix, the media processing device selects pixel points from the boundary pixel points in the currently obtained 8×8 dot matrix, and selects from the currently obtained 8×8 dot matrix. The pixels selected in the boundary pixels are operated to obtain an intermediate result corresponding to the NXN lattice composed of the currently obtained 8×8 lattice.
如果 N大于 8且不是 8的倍数， 那么媒体处理设备在得到对应 NXN点 阵的中间结果之后， 可以存储中间结果。媒体处理设备在得到对应 NXN点阵 的所有中间结果后， 可以对所有的中间结果进行运算，得到对应 NXN点阵的 像素点。 例如， 媒体处理设备可以对从当前获得的 8X8点阵中选取的像素点 进行求和运算， 求和运算的结果即为中间结果，媒体处理设备可以对所有的求 和运算的结果进行均值运算， 得到对应 NXN点阵的像素点， 其中， 均值运算 所使用的分母为所有参与运算的像素点的个数， 即， 当前获得的 8X8点阵参 与组成的 N X N点阵中所有参与运算的像素点的个数。 If N is greater than 8 and not a multiple of 8, then the media processing device is getting the corresponding NXN point After the intermediate results of the array, intermediate results can be stored. After obtaining all the intermediate results of the corresponding NXN lattice, the media processing device can calculate all the intermediate results to obtain the pixel points corresponding to the NXN lattice. For example, the media processing device may perform a summation operation on the pixels selected from the currently obtained 8×8 dot matrix, and the result of the summation operation is an intermediate result, and the media processing device may perform an average operation on the results of all the summation operations. Obtaining a pixel corresponding to the NXN lattice, wherein the denominator used in the mean operation is the number of pixels participating in the operation, that is, all the pixels participating in the NXN lattice that the currently obtained 8×8 lattice participates in. Number.
以 N=12为例， 媒体处理设备需要将接收到的图像信号缩小 12倍， 这种 情况下， 一个 8X8点阵需要输出 0.67X0.67个像素点， 但在实际应用中， 没 有非整数个数的像素点。 对此， 取 12和 8 的最小公倍数为 24, 这样， 3X3 个 8X8点阵输出 2X2个像素点。 在这个实施例中， 3X3个 8X8点阵可以被 划分为 2X2个 12X12点阵， 每个 12X12点阵都输出一个像素点。 如图 8所 示， 划分后的第 1个 12X12点阵由第 1个 8X8点阵、 第 2个 8X8点阵的第 1-4列、 第 4个 8X8点阵的第 1-4行和第 5个 8X8点阵的第 1-4行的第 1-4 列中的像素点组成， 其他 12X 12点阵的组成原理与此相同。 本领域技术人员 可以根据实际需要， 设计出参与第 1个 12X12点阵输出像素点的运算的像素 点， 使用平均值运算方式， 得到第 1个 12X12点阵输出的像素点的值。 同理， 其他 12X12点阵输出的像素点的值也可以通过运算得到。 需要说明的是， 图 8中的 8X8点阵所属的行、 8X8点阵之间的相邻关系以及接收 8X8点阵的顺 序可以参见上述 N=3的实施例中的相关描述， 这里不再赘述。 Taking N=12 as an example, the media processing device needs to reduce the received image signal by 12 times. In this case, an 8×8 dot matrix needs to output 0.67×0.67 pixels, but in practical applications, there is no non-integer number. The number of pixels. For this, take the least common multiple of 12 and 8 to 24, so that 3X3 8X8 dot matrix outputs 2X2 pixels. In this embodiment, 3X3 8X8 dots can be divided into 2X2 12X12 lattices, and each 12X12 dot matrix outputs one pixel. As shown in Fig. 8, the first 12X12 dot matrix after division is composed of the first 8×8 dot matrix, the second 8×8 dot matrix 1-4 column, the fourth 8×8 dot matrix row 1-4 and the The pixels in the 1-4th column of the 1-4th row of the 5 8X8 dot matrix are composed of the same, and the composition principle of the other 12X 12 dot matrix is the same. Those skilled in the art according to actual needs, the pixel design calculation involved in a 12X12 dot matrix of output pixel point, using the average value calculating mode, to obtain a value of the first pixel matrix of 12X12 output. Similarly, the values of the pixels of other 12X12 dot matrix outputs can also be obtained by operation. It should be noted that the sequence of the 8X8 dot matrix in FIG. 8 and the adjacent relationship between the 8×8 dot matrix and the order of receiving the 8×8 dot matrix can be referred to the related description in the above N=3 embodiment, and details are not described herein again. .
需要说明的是，由于媒体处理设备是按照先后顺序分别接收到图 8所示的 8X8点阵的， 所以， 如果第 1个 8X8点阵、 第 2个 8X8点阵的第 1-4列、 第 4个 8X8点阵的第 1-4行和第 5个 8X8点阵的第 1-4行的第 1-4列中均有像 素点需要参与输出像素点的运算， 那么媒体处理设备在获得第 1个 8X8点阵 后， 可以将第 1个 8X8点阵中参与输出像素点的运算的像素点的值相加， 緩 存结果， 在获得第 2个 8X8点阵后， 可以将第 2个 8X8点阵中参与第 1个 12X12点阵输出像素点的运算的像素点的值相加，緩存结果，在获得第 4个 8 X8点阵后， 可以将第 4个 8X8点阵中参与第 1个 12X 12点阵输出像素点的 运算的像素点的值相加， 緩存结果， 在获得第 5个 8X8点阵后， 可以将第 5 个 8X8点阵中参与第 1个 12X 12点阵输出像素点的运算的像素点的值相加， 得到一个结果， 最后， 将上述 4个结果做平均值运算， 得到第 1个 12X12点 阵输出的像素点的值， 其中， 平均值运算所使用的分母为第 1个 8X8点阵、 第 2个 8X8点阵、 第 4个 8X8点阵和第 5个 8X8点阵中所有参与运算的像 素点的个数。 其他 12X12点阵输出像素点的值的处理方式与此相同， 这里不 再赘述。 It should be noted that, since the media processing device receives the 8×8 dot matrix shown in FIG. 8 in sequence, therefore, if the first 8×8 dot matrix, the second 8×8 dot matrix, the first 1-4 column, the first In the 1-4th row of the 4th 8X8 dot matrix and the 1-4th row of the 1-4th row of the 5X8 dot matrix, there are pixels in the 1-4th column of the 8X8 dot matrix that need to participate in the operation of the output pixel point, then the media processing device obtains the After an 8X8 dot matrix, the values of the pixels participating in the operation of the output pixel in the first 8×8 dot matrix can be added, and the result can be cached. After the second 8×8 dot matrix is obtained, the second 8×8 dot can be obtained. The values of the pixels participating in the operation of the first 12X12 dot matrix output pixel are added, and the result is cached. After the fourth 8 X8 dot matrix is obtained, the fourth 8X8 dot matrix can participate in the first 12X. The value of the pixel point of the operation of the 12-dot output pixel is added, and the result is cached. After obtaining the 5th 8X8 dot matrix, the 5th bit can be obtained. The values of the pixels participating in the operation of the first 12×12 dot matrix output pixel in the 8×8 dot matrix are added to obtain a result. Finally, the above four results are averaged to obtain the first 12×12 dot matrix output. The value of the pixel, where the denominator used in the average operation is the pixel of the first 8×8 dot matrix, the second 8×8 dot matrix, the 4th 8×8 dot matrix, and the 5th 8×8 dot matrix. The number. The values of the other 12X12 dot matrix output pixel points are handled in the same manner, and will not be described here.
需要说明的是， 在上述实施例中， 如果 N大于 8且不是 8的倍数， 那么 对于涉及组成一个 NXN点阵的每个 8X8点阵都需要计算中间结果， 并对这 个 N X N点阵对应的所有中间结果进行运算，得到对应这个 N X N点阵的像素 点。在实际应用中，媒体处理设备也可以通过其他方式得到对应这个 N X N点 阵的像素点。 例如，媒体处理设备可以只从参与组成 NXN点阵的像素点数量 最多的 8X8点阵中选取像素点， 并对选取的像素点进行计算， 从而得到对应 NXN点阵的像素点， 简单的说， 可以认为参与组成 NXN点阵的像素点数量 最多的 8X8点阵代表了 NXN点阵。 还是以 N=12为例， 如图 8所示， 划分 后的第 1个 12X12点阵由第 1个 8X8点阵、 第 2个 8 X 8点阵的第 1-4列、 第 4个 8X8点阵的第 1-4行和第 5个 8X8点阵的第 1-4行的第 1-4列中的像 素点组成， 在计算对应 12X12点阵的像素点时， 由于第 1个 8X8点阵是参与 组成第 1个 12X12点阵像素点数量最多的 8X8点阵， 所以， 可以只从第 1 个 8X8点阵中选取像素点， 并对选取的像素点进行运算， 从而得到对应第 1 个 12X12点阵的像素点。 为使本领域技术人员更加清楚的理解本发明，下面再结合图 9对本发明进 行说明。 如图 9所示， 本发明的实现流程可以包括：  It should be noted that, in the above embodiment, if N is greater than 8 and not a multiple of 8, then intermediate results need to be calculated for each 8×8 dot matrix involved in forming an NXN lattice, and all corresponding to the NXN lattice The intermediate result is calculated to obtain the pixel points corresponding to the NXN lattice. In practical applications, the media processing device may also obtain pixel points corresponding to the N X N dot matrix by other means. For example, the media processing device may select pixel points only from the 8×8 dot matrix that participates in the largest number of pixels that make up the NXN dot matrix, and calculate the selected pixel points to obtain pixel points corresponding to the NXN dot matrix. In short, It can be considered that the 8X8 dot matrix that participates in the largest number of pixels that make up the NXN lattice represents the NXN dot matrix. Or take N=12 as an example. As shown in Fig. 8, the first 12X12 dot matrix after division is composed of the first 8X8 dot matrix, the second 8×8 dot matrix, the 1-4th column, and the fourth 8X8. The pixels in the 1-4th row of the dot matrix and the 1-4th row in the 1-4th row of the 5th 8X8 dot matrix are composed of the pixel points corresponding to the 12X12 dot matrix, due to the first 8X8 dot. The array is the 8X8 dot matrix that constitutes the largest number of pixels in the first 12X12 dot matrix. Therefore, pixels can be selected only from the first 8×8 dot matrix, and the selected pixel points are calculated to obtain the corresponding first pixel. The pixel of the 12X12 dot matrix. In order to make the present invention more clearly understood by those skilled in the art, the present invention will be described below in conjunction with FIG. As shown in FIG. 9, the implementation process of the present invention may include:
S901: 获得缩小倍数 N;  S901: obtaining a reduction factor N;
S902: 判断 N是否小于 8, 如果是， 则转 S903, 否则， 转 S908;  S902: Determine whether N is less than 8, if yes, then go to S903, otherwise, go to S908;
S903: 判断 N是否为 8的约数， 如果是， 则转 S904, 否则， 转 S906; S904: 选择对应 NXN点阵的基本模板， 转 S905;  S903: Determine whether N is a divisor of 8, if yes, then go to S904, otherwise, go to S906; S904: select the basic template corresponding to the NXN lattice, and then go to S905;
S905: 根据选择的基本模板， 分别从 8X8点阵中的每个 NXN点阵中选 取像素点， 并分别对选取的像素点进行运算， 分别得到与 8X8点阵中的每个 N X N点阵对应的像素点， 转 S913； S905: Select, according to the selected basic template, pixel points from each of the NXN lattices in the 8×8 dot matrix, and perform operations on the selected pixel points respectively, respectively obtaining each of the 8×8 dot matrixes. The pixel corresponding to the NXN dot matrix, turn to S913;
S906: 选择对应 NXN点阵的基本模板和对应边界像素点的基本模板， 转 S907;  S906: Select a basic template corresponding to the NXN lattice and a basic template corresponding to the boundary pixel, and then go to S907;
S907: 根据选择的基本模板， 分别从 8X8点阵中的每个 NXN点阵和边 界像素点中选取像素点， 并分别对选择的像素点进行运算， 分别得到与 8X8 点阵中的每个 NXN点阵和边界像素点对应的像素点， 转 S913;  S907: Select, according to the selected basic template, pixel points from each of the NXN dot matrix and the boundary pixel in the 8×8 dot matrix, and perform operations on the selected pixel points respectively, respectively obtaining each NXN in the 8×8 dot matrix. The pixel corresponding to the dot matrix and the boundary pixel point, and then go to S913;
S908: 判断 N是否为 8的倍数， 如果是， 则转 S909, 否则， 转 S911; S909： 选择对应 8 X 8点阵的基本模板， 转 S910；  S908: Determine whether N is a multiple of 8, if yes, then go to S909, otherwise, go to S911; S909: select the basic template corresponding to 8 X 8 dot matrix, and turn to S910;
S910: 根据选择的基本模板， 分别从参与组成 NXN点阵的每个 8X8点 阵中选取像素点，并分别对选取的像素点进行运算，得到多个中间结果，其中， 参与组成 NXN点阵的每个 8X8点阵都对应一个中间结果， 之后， 对所有的 中间结果进行运算， 得到对应 NXN点阵的像素点， 转 S913;  S910: Select, according to the selected basic template, pixel points from each 8×8 dot matrix participating in the NXN dot matrix, and perform operations on the selected pixel points respectively to obtain multiple intermediate results, where the components participating in the NXN dot matrix are formed. Each 8X8 dot matrix corresponds to an intermediate result, after which all the intermediate results are calculated to obtain the pixel points corresponding to the NXN dot matrix, and the process proceeds to S913;
S911: 选择对应参与组成 N X N点阵的 8 X 8点阵中的所有像素点和 8 X 8 点阵中的边界像素点的基本模板， 转 S912;  S911: selecting a basic template corresponding to all pixel points in the 8×8 dot matrix that constitutes the N X N dot matrix and boundary pixel points in the 8×8 dot matrix, and then turning to S912;
S912: 根据选择的基本模板， 分别从参与组成 NXN点阵的每个 8X8点 阵中选取像素点，并分别对选取的像素点进行运算，得到多个中间结果，其中， 参与组成 NXN点阵的每个 8X8点阵都对应一个中间结果， 之后， 对所有的 中间结果进行运算， 得到对应 NXN点阵的像素点， 转 S913;  S912: Select, according to the selected basic template, pixel points from each 8×8 dot matrix participating in the NXN dot matrix, and perform operations on the selected pixel points respectively to obtain a plurality of intermediate results, wherein, the components participating in the NXN dot matrix are formed. Each 8X8 dot matrix corresponds to an intermediate result, after which all the intermediate results are calculated to obtain the pixel points corresponding to the NXN dot matrix, and the process proceeds to S913;
S913: 输出得到的像素点。 对应于图 3 所示的方法实施例， 本发明还提供一种媒体处理设备。 如图 10所示， 这种媒体处理设备包括： 获得单元 1001, 用于获得以 8X8点阵为单 位的图像信号； 选取单元 1002, 用于从当前获得的 8X8点阵中选取像素点， 选取的像素点用于参与将以 8X8点阵为单位的图像信号缩小 N倍的处理， 其 中， N为正整数； 运算单元 1003, 用于对从当前获得的 8X8点阵中选取的像 素点进行运算； 如果 N小于 8, 则得到对应 NXN点阵的像素点； 如果 N大 于 8, 则得到对应 NXN点阵的中间结果； 其中， 对于每个 NXN点阵都得到 一个像素点， 如果 N大于 8, 则对于每个 NXN点阵得到的像素点由多个中间 结果进行运算得到。 前面提到过， 媒体处理设备是以 8X8点阵为单位接收图 像的， 所以， 获得单元 1001获得的是组成图像的 8X8点阵。 媒体处理设备获 得 8X8点阵后， 可以緩存 8X8点阵， 所以， 媒体处理设备还可以包括一个具 有緩存功能的緩存单元（图 10中未绘示 ), 获得单元 1001获得 8X8点阵后， 可以将 8X8点阵发给緩存单元进行緩存。选取单元 1002可以从緩存单元緩存 的 8X8点阵中选取参与运算的像素点。 S913: Output the obtained pixel point. Corresponding to the method embodiment shown in FIG. 3, the present invention also provides a media processing device. As shown in FIG. 10, the media processing device includes: an obtaining unit 1001, configured to obtain an image signal in units of 8×8 dot matrix; and a selecting unit 1002, configured to select a pixel point from the currently obtained 8×8 dot matrix, and selected The pixel is used to participate in a process of reducing the image signal in units of 8×8 dot matrix by N times, wherein N is a positive integer; and the operation unit 1003 is configured to perform operation on the pixel selected from the currently obtained 8×8 dot matrix; If N is less than 8, the pixel corresponding to the NXN lattice is obtained; if N is greater than 8, an intermediate result corresponding to the NXN lattice is obtained; wherein, for each NXN lattice, one pixel is obtained, and if N is greater than 8, The pixels obtained for each NXN lattice are computed from a plurality of intermediate results. As mentioned earlier, the media processing device receives the image in units of 8×8 dots. Image, so, the obtaining unit 1001 obtains an 8×8 dot matrix that constitutes an image. After the 8X8 dot matrix is obtained, the media processing device can cache the 8X8 dot matrix. Therefore, the media processing device can also include a buffering unit with a buffer function (not shown in FIG. 10). After obtaining the 8×8 dot matrix, the obtaining unit 1001 can The 8X8 dot matrix is sent to the cache unit for caching. The selecting unit 1002 can select the pixel points participating in the operation from the 8×8 dot matrix buffered by the buffer unit.
上述的媒体处理设备还可以包括存储单元 1004, 用于存储运算单元 1003 得到的中间结果。 运算单元 1003在得到组成 NXN点阵的所有中间结果后， 可以对所有的中间结果进行运算， 得到对应 NXN点阵的像素点。  The media processing device described above may further include a storage unit 1004 for storing intermediate results obtained by the arithmetic unit 1003. After obtaining all the intermediate results constituting the NXN lattice, the arithmetic unit 1003 can perform all the intermediate results to obtain the pixel points corresponding to the NXN lattice.
运算单元 1003具体可以用于对选取单元 1002选取的像素点进行均值运 算。  The operation unit 1003 is specifically configured to perform an average operation on the pixels selected by the selection unit 1002.
运算单元 1003具体可以用于对选取单元 1002从获得单元 1001当前获得 的 8X8点阵中选取的像素点进行求和运算。 对应 NXN点阵的中间结果即为 运算单元 1003对选取单元 1002从获得单元 1001当前获得的 8X8点阵中选取 的像素点进行求和运算的结果。 运算单元 1003具体可以用于对与 NXN点阵 对应的所有求和运算的结果进行均值运算，得到对应 NXN点阵的像素点， 其 中， 均值运算所使用的分母为所有参与运算的像素点的个数。  The operation unit 1003 may be specifically configured to perform a summation operation on the pixel points selected by the selection unit 1002 from the 8×8 dot matrix currently obtained by the obtaining unit 1001. The intermediate result corresponding to the NXN lattice is the result of the summation operation performed by the operation unit 1003 on the pixel points selected by the selection unit 1002 from the 8X8 dot matrix currently obtained by the obtaining unit 1001. The operation unit 1003 may be specifically configured to perform a mean operation on the results of all the summation operations corresponding to the NXN lattice to obtain pixel points corresponding to the NXN lattice, wherein the denominator used in the mean operation is all the pixels participating in the operation. number.
由于上述的媒体处理设备的实施例与图 3 所示的方法实施例相对应， 所 以， 上述的媒体处理设备中的各个单元的功能以及相互配合关系可以参见图 3 所示的方法实施例中的相关描述， 这里不再赘述。 另外， 本发明还提供一种图像处理***， 包括解码器、 显示设备和上述的 媒体处理设备。媒体处理设备的功能以及媒体处理设备与解码器、显示设备之 间的配合关系可以参见上文的描述， 这里不再赘述。 综上所述， 在本发明中， 媒体处理设备获得以 8X8点阵为单位的图像信 号后，从当前获得的 8X8点阵中选取用于参与将以 8X8点阵为单位的图像信 号缩小 N倍的处理的像素点， 对选取的像素点进行运算， 输出对应 NXN点 阵的像素点或中间结果， 无须緩存多个 8X8点阵后， 再进行缩小处理， 节省 了为緩存多个 8X8点阵而需要的緩存资源。 另外， 在本发明中， 媒体处理设备对从当前获得的 8 X 8点阵中选取的像 素点进行运算， 得到像素点或者中间结果， 不需要进行抗混叠滤波， 实现较为 简捷。 For the above-mentioned embodiments of the media processing device and the method embodiment shown in FIG. 3, the function and the mutual cooperation relationship of each unit in the foregoing media processing device can be referred to the method embodiment shown in FIG. Related descriptions are not described here. In addition, the present invention also provides an image processing system including a decoder, a display device, and the above-described media processing device. The function of the media processing device and the cooperation relationship between the media processing device and the decoder and the display device can be referred to the above description, and details are not described herein again. In summary, in the present invention, after obtaining the image signal in units of 8×8 dot matrix, the media processing device selects from the currently obtained 8×8 dot matrix to participate in reducing the image signal in units of 8×8 dot matrix by N times. The processed pixel points, perform operations on the selected pixel points, output pixel points or intermediate results corresponding to the NXN dot matrix, and do not need to cache multiple 8×8 dot matrixes, and then perform reduction processing, thereby saving a plurality of 8×8 dot matrixes for buffering. The required cache resources. In addition, in the present invention, the media processing device performs calculation on the pixel points selected from the currently obtained 8×8 dot matrix to obtain pixel points or intermediate results, and does not need anti-aliasing filtering, and the implementation is relatively simple.
通过以上的实施方式的描述，本领域的技术人员可以清楚地了解到本发明 可以通过硬件实现， 也可以可借助软件加必要的通用硬件平台的方式来实现， 基于这样的理解， 本发明的技术方案可以以软件产品的形式体现出来， 该软件 产品可以存储在一个非易失性存储介质 （可以是 CD-ROM, U盘， 移动硬盘 等）中， 包括若干指令用以使得一台计算机设备 (可以是个人计算机，服务器， 或者网络设备等）执行本发明各个实施例所述的方法。  Through the description of the above embodiments, those skilled in the art can clearly understand that the present invention can be implemented by hardware, or can be implemented by means of software plus a necessary general hardware platform. Based on such understanding, the technology of the present invention. The solution can be embodied in the form of a software product that can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including a number of instructions for making a computer device ( The method described in various embodiments of the present invention may be a personal computer, a server, or a network device.
总之， 以上所述仅为本发明的较佳实施例而已， 并非用于限定本发明的保 护范围。凡在本发明的精神和原则之内，所作的任何修改、等同替换、改进等， 均应包含在本发明的保护范围之内。  In summary, the above description is only a preferred embodiment of the present invention and is not intended to limit the scope of the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and scope of the invention are intended to be included within the scope of the invention.

Claims

权 利 要 求 Rights request
1.一种图像缩小方法， 其特征在于， 适用于图像预览模式， 应用于包括解 码器和媒体处理设备的图像处理***， 媒体处理设备对解码器输出的以 8X8 点阵为单位的图像信号进行缩小处理， 所述以 8X8点阵为单位的图像信号是 解码器经过离散余弦变换 DCT解码后得到的图像信号， 所述方法包括：  An image reduction method, which is suitable for an image preview mode, applied to an image processing system including a decoder and a media processing device, wherein the media processing device performs an image signal output by the decoder in an 8×8 dot matrix unit. The image processing signal in the 8×8 dot matrix is an image signal obtained by the decoder after the discrete cosine transform DCT decoding, and the method includes:
媒体处理设备获得以 8X8点阵为单位的图像信号；  The media processing device obtains an image signal in units of 8×8 dots;
媒体处理设备从当前获得的 8X8点阵中选取像素点， 用于参与将以 8X8 点阵为单位的图像信号缩小 N倍的处理， 其中， N为正整数；  The media processing device selects a pixel from the currently obtained 8×8 dot matrix for participating in a process of reducing the image signal in units of 8×8 dot matrix by N times, wherein N is a positive integer;
媒体处理设备对从当前获得的 8X8点阵中选取的像素点进行运算； 如果 N小于 8, 则得到对应 NXN点阵的像素点；  The media processing device calculates a pixel selected from the currently obtained 8×8 dot matrix; if N is less than 8, the pixel corresponding to the NXN dot matrix is obtained;
如果 N大于 8, 则得到对应 NXN点阵的中间结果；  If N is greater than 8, an intermediate result corresponding to the NXN lattice is obtained;
其中， 对于每个 NXN点阵都得到一个像素点， 如果 N大于 8, 则对于每 个 NXN点阵得到的像素点由多个中间结果进行运算得到。  Among them, one pixel is obtained for each NXN lattice. If N is greater than 8, the pixel obtained for each NXN lattice is calculated by multiple intermediate results.
2.如权利要求 1所述的方法， 其特征在于， 如果 N是 8的约数， 则当前获 得的 8X8点阵被划分为 （8/N) X (8/N)个 NXN点阵。  The method according to claim 1, wherein if N is a divisor of 8, the currently obtained 8X8 lattice is divided into (8/N) X (8/N) NXN lattices.
3.如权利要求 1所述的方法，其特征在于，如果 N小于 8且不是 8的约数， 则当前获得的 8X8点阵被划分为至少一个 NXN点阵和多个边界像素点， 当 前获得的 8 X 8点阵中的边界像素点与相邻的 8 X 8点阵中的边界像素点组成 N XN点阵。  3. The method of claim 1 wherein if N is less than 8 and not a divisor of 8, the currently obtained 8X8 lattice is divided into at least one NXN lattice and a plurality of boundary pixel points, currently obtained The boundary pixel points in the 8 X 8 dot matrix and the boundary pixel points in the adjacent 8 X 8 dot matrix form an N XN dot matrix.
4.如权利要求 2或 3所述的方法， 其特征在于， 媒体处理设备从当前获得 的 8X8点阵中选取像素点具体为：媒体处理设备分别从当前获得的 8X8点阵 中的每个 N X N点阵中选取像素点；  The method according to claim 2 or 3, wherein the media processing device selects pixel points from the currently obtained 8×8 dot matrix, specifically: the media processing device respectively obtains each NXN in the currently obtained 8×8 dot matrix. Select a pixel in the dot matrix;
媒体处理设备对从当前获得的 8X8 点阵中选取的像素点进行运算具体 为： 媒体处理设备分别对从当前获得的 8X8点阵中的每个 NXN点阵中选取 的像素点进行运算；  The media processing device performs operations on the pixels selected from the currently obtained 8×8 dot matrix as follows: The media processing device respectively calculates the pixels selected from each of the currently selected 8×8 dot matrixes;
媒体处理设备得到对应 NXN点阵的像素点具体为：媒体处理设备分别得 到对应当前获得的 8X8点阵中的每个 NXN点阵的像素点。  The media processing device obtains the pixel points corresponding to the NXN dot matrix, specifically: the media processing device respectively obtains the pixel points corresponding to each of the currently selected 8×8 dot matrix.
5.如权利要求 3所述的方法， 其特征在于， 媒体处理设备按照下述方式从 当前获得的 8X8点阵中选取像素点：如果当前获得的 8X8点阵中用于组成一 个 NXN点阵的边界像素点的数量比相邻的 8X8点阵中用于组成所述 NXN 点阵的边界像素点的数量多， 则媒体处理设备从当前获得的 8X8点阵中用于 组成所述 N X N点阵的边界像素点中选取像素点； The method according to claim 3, wherein the media processing device selects pixel points from the currently obtained 8×8 dot matrix in the following manner: if the currently obtained 8×8 dot matrix is used to form one The number of boundary pixels of the NXN lattice is larger than the number of boundary pixels used to form the NXN lattice in the adjacent 8×8 dot matrix, and the media processing device is used for the composition from the currently obtained 8×8 dot matrix. Selecting a pixel point from the boundary pixel of the NXN dot matrix;
媒体处理设备对从当前获得的 8X8 点阵中选取的像素点进行运算具体 为： 媒体处理设备对从当前获得的 8X8点阵中用于组成所述 NXN点阵的边 界像素点中选取的像素点进行运算；  The media processing device performs operations on the pixels selected from the currently obtained 8×8 dot matrix, specifically: the media processing device selects the pixel points selected from the currently selected 8×8 dot matrix for the boundary pixel points constituting the NXN dot matrix. Perform an operation;
媒体处理设备得到对应 NXN点阵的像素点具体为：媒体处理设备得到对 应所述 N X N点阵的像素点。  The pixel of the media processing device corresponding to the NXN dot matrix is specifically: the media processing device obtains a pixel corresponding to the N X N dot matrix.
6.如权利要求 1所述的方法， 其特征在于， 如果 N为 8的倍数， 则 kXk 个 8X8点阵组成 NXN点阵， 其中， k=N/8。  The method according to claim 1, wherein if N is a multiple of 8, kXk 8X8 lattices constitute an NXN lattice, wherein k = N/8.
7.如权利要求 6所述的方法， 其特征在于， 媒体处理设备得到对应 NX N 点阵的中间结果具体为： 媒体处理设备得到中间结果， 所述中间结果与当前获 得的 8 X 8点阵参与组成的 N X N点阵相对应。  The method according to claim 6, wherein the media processing device obtains an intermediate result corresponding to the NX N lattice, specifically: the media processing device obtains an intermediate result, and the intermediate result and the currently obtained 8×8 dot matrix Participate in the composition of the NXN lattice.
8.如权利要求 1所述的方法，其特征在于，如果 N大于 8且不是 8的倍数， 则 NXN点阵由至少一个 8X8点阵和相邻的 8X8点阵中的边界像素点组成。  8. The method of claim 1 wherein if N is greater than 8 and not a multiple of 8, the NXN lattice is comprised of at least one 8X8 lattice and boundary pixel points in an adjacent 8X8 lattice.
9.如权利要求 8所述的方法， 其特征在于， 媒体处理设备从当前获得的 8 X8点阵中选取像素点具体为： 如果当前获得的 8X8点阵中的所有像素点都 参与组成 NXN点阵， 则媒体处理设备从当前获得的 8X8点阵中的所有像素 点中选取像素点； 如果当前获得的 8X8点阵中的边界像素点参与组成 NXN 点阵，则媒体处理设备从当前获得的 8X8点阵中的边界像素点中选取像素点； 媒体处理设备对从当前获得的 8X8 点阵中选取的像素点进行运算具体 为： 如果当前获得的 8X8点阵中的所有像素点都参与组成 NXN点阵， 则媒 体处理设备对从当前获得的 8X8点阵中的所有像素点中选取的像素点进行运 算； 如果当前获得的 8X8点阵中的边界像素点参与组成 NXN点阵， 则媒体 处理设备对从当前获得的 8X8 点阵中的边界像素点中选取的像素点进行运 算；  The method according to claim 8, wherein the media processing device selects pixel points from the currently obtained 8×8 dot matrix, specifically: if all pixels in the currently obtained 8×8 dot matrix participate in forming the NXN point Array, the media processing device selects pixel points from all the pixels in the currently obtained 8×8 dot matrix; if the boundary pixel points in the currently obtained 8×8 dot matrix participate in forming the NXN dot matrix, the media processing device obtains the currently obtained 8×8 The pixel points are selected from the boundary pixels in the dot matrix; the media processing device performs operations on the pixels selected from the currently obtained 8×8 dot matrix as follows: If all the pixels in the currently obtained 8×8 dot matrix participate in forming the NXN point Array, the media processing device calculates the pixel points selected from all the pixels in the currently obtained 8×8 dot matrix; if the boundary pixel points in the currently obtained 8×8 dot matrix participate in the NXN dot matrix, the media processing device pairs Calculate from the pixel points selected from the boundary pixels in the currently obtained 8X8 dot matrix;
媒体处理设备得到对应 NXN点阵的中间结果具体为：媒体处理设备得到 与当前获得的 8 X 8点阵参与组成的 N X N点阵对应的中间结果。  The intermediate result of the media processing device corresponding to the NXN lattice is specifically as follows: the media processing device obtains an intermediate result corresponding to the N X N dot matrix composed of the currently obtained 8 X 8 dot matrix.
10.如权利要求 1、 6-9任意一项所述的方法， 其特征在于， 媒体处理设备 得到对应 N X N点阵的中间结果之后， 存储所述中间结果； The method according to any one of claims 1 to 6, wherein the media processing device After obtaining an intermediate result corresponding to the NXN lattice, storing the intermediate result;
媒体处理设备在得到对应 NXN点阵的所有中间结果后，对所有的中间结 果进行运算， 得到对应 NXN点阵的像素点。  After obtaining all the intermediate results of the corresponding NXN lattice, the media processing device calculates all the intermediate results to obtain the pixels corresponding to the NXN lattice.
11.如权利要求 1-3、 5-9任意一项所述的方法， 其特征在于， 媒体处理设 备对从当前获得的 8 X 8点阵中选取的像素点进行运算具体为： 媒体处理设备 对选取的像素点进行均值运算。  The method according to any one of claims 1-3, 5-9, wherein the media processing device performs operations on pixels selected from the currently obtained 8×8 dot matrix, specifically: media processing device Perform the mean operation on the selected pixels.
12.如权利要求 10所述的方法， 其特征在于， 媒体处理设备对从当前获得 的 8 X 8点阵中选取的像素点进行运算具体为： 媒体处理设备对从当前获得的 8 X 8点阵中选取的像素点进行求和运算；  The method according to claim 10, wherein the media processing device performs operations on the pixels selected from the currently obtained 8×8 dot matrix, specifically: the media processing device pairs the currently obtained 8×8 points. The pixels selected in the array are summed;
所述对应 NXN点阵的中间结果为媒体处理设备对从当前获得的 8 X 8点 阵中选取的像素点进行求和运算的结果；  The intermediate result of the corresponding NXN dot matrix is a result of the summation operation performed by the media processing device on the pixel points selected from the currently obtained 8×8 dot matrix;
媒体处理设备对所有的中间结果进行运算、得到对应 N X N点阵的像素点 具体为： 媒体处理设备对所有的求和运算的结果进行均值运算，得到对应 NX N点阵的像素点， 其中， 均值运算所使用的分母为所有参与运算的像素点的个 数。  The media processing device calculates all the intermediate results and obtains the pixel points corresponding to the NXN dot matrix. The media processing device performs the mean operation on the results of all the summation operations to obtain the pixel points corresponding to the NX N dot matrix, wherein, the mean value The denominator used in the operation is the number of pixels participating in the operation.
13.—种媒体处理设备， 其特征在于， 适用于图像预览模式， 应用于包括 解码器和媒体处理设备的图像处理*** , 媒体处理设备对解码器输出的以 8 X 8点阵为单位的图像信号进行缩小处理， 所述以 8 X 8点阵为单位的图像信号 是解码器经过离散余弦变换 DCT解码后得到的图像信号， 所述媒体处理设备 包括：  13. A media processing device, characterized in that it is suitable for an image preview mode, applied to an image processing system including a decoder and a media processing device, and an image of the 8 X 8 dot matrix output by the media processing device to the decoder The signal processing is performed, and the image signal in the unit of 8×8 dot matrix is an image signal obtained by the decoder after the discrete cosine transform DCT decoding, and the media processing device includes:
获得单元， 用于获得以 8 X 8点阵为单位的图像信号；  Obtaining unit for obtaining an image signal in units of 8×8 dots;
选取单元， 用于从当前获得的 8 X 8点阵中选取像素点， 选取的像素点用 于参与将以 8 X 8点阵为单位的图像信号缩小 N倍的处理，其中， N为正整数； 运算单元， 用于对从当前获得的 8 X 8点阵中选取的像素点进行运算； 如 果 N小于 8, 则得到对应 NXN点阵的像素点； 如果 N大于 8, 则得到对应 N XN点阵的中间结果；  a selecting unit, configured to select a pixel from the currently obtained 8×8 dot matrix, and the selected pixel is used to participate in a process of reducing an image signal of 8×8 dot matrix by N times, wherein N is a positive integer The operation unit is configured to perform calculation on the pixel selected from the currently obtained 8×8 dot matrix; if N is less than 8, the pixel point corresponding to the NXN lattice is obtained; if N is greater than 8, the corresponding N XN point is obtained Intermediate result of the array;
其中， 对于每个 NXN点阵都得到一个像素点， 如果 N大于 8, 则对于每 个 NXN点阵得到的像素点由多个中间结果进行运算得到。  Among them, one pixel is obtained for each NXN lattice. If N is greater than 8, the pixel obtained for each NXN lattice is calculated by multiple intermediate results.
14.如权利要求 13所述的媒体处理设备， 其特征在于，还包括： 存储单元， 用于存储所述运算单元得到的中间结果； The media processing device according to claim 13, further comprising: a storage unit, An intermediate result obtained by storing the arithmetic unit;
所述运算单元在得到组成 NXN点阵的所有中间结果后，对所有的中间结 果进行运算， 得到对应 NXN点阵的像素点。  After obtaining all the intermediate results constituting the NXN lattice, the arithmetic unit performs operations on all the intermediate results to obtain pixel points corresponding to the NXN lattice.
15.如权利要求 13所述的媒体处理设备， 其特征在于， 所述运算单元具体 用于对所述选取单元选取的像素点进行均值运算。  The media processing device according to claim 13, wherein the operation unit is specifically configured to perform an average operation on the pixel points selected by the selection unit.
16.如权利要求 14所述的媒体处理设备， 其特征在于， 所述运算单元具体 用于对所述选取单元从所述获得单元当前获得的 8 X 8点阵中选取的像素点进 行求和运算；  The media processing device according to claim 14, wherein the operation unit is specifically configured to: sum the pixel points selected by the selection unit from the 8×8 dot matrix currently obtained by the obtaining unit Operation
所述对应 NXN 点阵的中间结果为所述运算单元对所述选取单元从所述 获得单元当前获得的 8 X 8点阵中选取的像素点进行求和运算的结果；  The intermediate result of the corresponding NXN dot matrix is a result of the summation operation performed by the operation unit on the pixel points selected by the selection unit from the 8×8 dot matrix currently obtained by the obtaining unit;
所述运算单元具体用于对与 NX N点阵对应的所有求和运算的结果进行 均值运算， 得到对应 NXN点阵的像素点， 其中， 均值运算所使用的分母为所 有参与运算的像素点的个数。  The operation unit is specifically configured to perform a mean operation on the results of all the summation operations corresponding to the NX N lattice to obtain a pixel point corresponding to the NXN lattice, wherein the denominator used in the mean operation is all the pixels participating in the operation. Number.
17.—种图像处理***， 其特征在于， 包括解码器、 显示设备和权利要求 13-16任意一项所述的媒体处理设备。  17. An image processing system, comprising: a decoder, a display device, and the media processing device of any of claims 13-16.
PCT/CN2010/072270 2010-04-28 2010-04-28 Image downscaling method and apparatus WO2011134144A1 (en)

Priority Applications (2)

Application Number Priority Date Filing Date Title
PCT/CN2010/072270 WO2011134144A1 (en) 2010-04-28 2010-04-28 Image downscaling method and apparatus
CN201080061138.2A CN102893592B (en) 2010-04-28 2010-04-28 Image downscaling method and device

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2010/072270 WO2011134144A1 (en) 2010-04-28 2010-04-28 Image downscaling method and apparatus

Publications (1)

Publication Number Publication Date
WO2011134144A1 true WO2011134144A1 (en) 2011-11-03

Family

ID=44860757

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2010/072270 WO2011134144A1 (en) 2010-04-28 2010-04-28 Image downscaling method and apparatus

Country Status (2)

Country Link
CN (1) CN102893592B (en)
WO (1) WO2011134144A1 (en)

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1414793A (en) * 2001-10-23 2003-04-30 三星电子株式会社 Compression video decoder with contraction image function and its method
CN1523894A (en) * 2002-12-11 2004-08-25 精工爱普生株式会社 Image scaling in the compressed domain
US20050175258A1 (en) * 2004-02-11 2005-08-11 Leung Patrick W. Method and apparatus for downscaling digital image data to fit a graphics display device
CN101483036A (en) * 2008-01-10 2009-07-15 扬智科技股份有限公司 Apparatus for image reduction and method thereof
CN101692336A (en) * 2009-10-09 2010-04-07 青岛海信宽带多媒体技术有限公司 Method for zooming out images

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1414793A (en) * 2001-10-23 2003-04-30 三星电子株式会社 Compression video decoder with contraction image function and its method
CN1523894A (en) * 2002-12-11 2004-08-25 精工爱普生株式会社 Image scaling in the compressed domain
US20050175258A1 (en) * 2004-02-11 2005-08-11 Leung Patrick W. Method and apparatus for downscaling digital image data to fit a graphics display device
CN101483036A (en) * 2008-01-10 2009-07-15 扬智科技股份有限公司 Apparatus for image reduction and method thereof
CN101692336A (en) * 2009-10-09 2010-04-07 青岛海信宽带多媒体技术有限公司 Method for zooming out images

Also Published As

Publication number Publication date
CN102893592B (en) 2015-09-30
CN102893592A (en) 2013-01-23

Similar Documents

Publication Publication Date Title
CN110136066B (en) Video-oriented super-resolution method, device, equipment and storage medium
TWI299838B (en) Image processing apparatus and image processing method
TW525391B (en) Method and apparatus for performing motion compensation in a texture mapping engine
JP2019118101A (en) Video coding method and image processing apparatus therefor
JP2004274724A (en) Method and device for reconfiguring high-resolution image
WO2011134144A1 (en) Image downscaling method and apparatus
CN101778280A (en) Circuit and method based on AVS motion compensation interpolation
CN101771867A (en) Size-reducing decoding method and system
TWI364222B (en) Method and device for generating prediction mode parameter
JP2888292B2 (en) Two-dimensional inverse discrete cosine transform (IDCT) device
US20210335018A1 (en) Data generating device, training device, and data generating method
JP2946612B2 (en) IDCT processing equipment
TW201322774A (en) Multiple stream processing for video analytics and encoding
JP4740992B2 (en) Method and apparatus for performing overlap filtering and core conversion
JP4266512B2 (en) Data processing device
JP6490272B2 (en) Image encoding device, image encoding method and program, image decoding device, image decoding method and program
JP4700838B2 (en) Filter processing device
JP4217408B2 (en) Filter processing device
KR101383515B1 (en) Apparatus for inverse-transforming image
CN101783857B (en) Image matrixing pretreatment method based on FPGA in high-resolution imaging system
JP3302531B2 (en) Line conversion circuit of image processing device
US6560286B1 (en) Field frame motion design for digital video decoder
JPH06274607A (en) Parallel signal processor
KR100836534B1 (en) A low power sub-pixel interpolator architecture
JP2010154001A (en) Image compression processing unit and method of compressing image

Legal Events

Date Code Title Description
WWE Wipo information: entry into national phase

Ref document number: 201080061138.2

Country of ref document: CN

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

Ref document number: 10850477

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: 10850477

Country of ref document: EP

Kind code of ref document: A1