WO2010001911A1 - フィルタ装置 - Google Patents
フィルタ装置 Download PDFInfo
- Publication number
- WO2010001911A1 WO2010001911A1 PCT/JP2009/062003 JP2009062003W WO2010001911A1 WO 2010001911 A1 WO2010001911 A1 WO 2010001911A1 JP 2009062003 W JP2009062003 W JP 2009062003W WO 2010001911 A1 WO2010001911 A1 WO 2010001911A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- pixel
- value
- adjacent
- block
- pixel value
- Prior art date
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/85—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using pre-processing or post-processing specially adapted for video compression
- H04N19/86—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using pre-processing or post-processing specially adapted for video compression involving reduction of coding artifacts, e.g. of blockiness
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/102—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or selection affected or controlled by the adaptive coding
- H04N19/117—Filters, e.g. for pre-processing or post-processing
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/134—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or criterion affecting or controlling the adaptive coding
- H04N19/136—Incoming video signal characteristics or properties
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/169—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding
- H04N19/17—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object
- H04N19/176—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object the region being a block, e.g. a macroblock
Definitions
- the present invention relates to a filter device that removes block noise generated when image information encoded by dividing into blocks of different sizes is decoded.
- Image coding technology is applied to many familiar video devices, from television receivers to mobile phones capable of image processing.
- image data (image information) is divided into a plurality of blocks, orthogonal transform is performed in units of the divided blocks, and the obtained transform coefficients are quantized and then quantized.
- the converted transform coefficient is encoded by variable length encoding.
- the image is degraded due to loss of information in the quantization.
- a large distortion is likely to occur in an image due to a large distortion (so-called block noise) generated at the boundary of each block which is a unit for performing orthogonal transformation.
- block noise occurs at the boundary of each block in the image encoding process, block noise also occurs when the encoded image is decoded, so that the user viewing the image can easily feel unnaturalness. Therefore, a general image decoding device (or image encoding device) uses a filter device that filters block noise generated at the boundary of each block in order to remove block noise.
- Non-Patent Document 1 discloses a basic filter processing technique. The filter processing technique described in Non-Patent Document 1 will be described below with reference to FIG.
- FIG. 16 is a diagram schematically illustrating a pixel at a block boundary and a pixel value of the pixel.
- a pixel boundary between the pixel P0 having the pixel value p0 and the pixel Q0 having the pixel value q0 is a block boundary, and as the distance from the boundary increases, the pixels P1, P2, P3, and the pixels Q1, Q2, Q3.
- pixel values in each pixel are set to pixel values p1, p2, and p3 and pixel values q1, q2, and q3, respectively. That is, in FIG. 16, among two adjacent blocks at the block boundary, the pixel value of one block is the pixel value pk (k is a value determined by the distance from the boundary), and the pixel value of the other block is the pixel value qk. (K is a value determined by the distance from the boundary). In FIG. 16, it is not distinguished whether the boundary is a horizontal boundary or a vertical boundary.
- the filter device performs a filtering process when the calculated “d” and “ap” are d ⁇ and ap ⁇ with respect to the predetermined threshold values ⁇ and ⁇ . Note that if the above condition is not satisfied, the filter processing for the processing target pixel is not executed.
- the filter device is a processing target of the filter process based on the filter coefficients (weighting) shown in FIGS. 17A and 17B.
- FIGS. 17A and 17B are diagrams showing filter coefficients at each pixel value when the filter process is performed.
- Non-Patent Document 1 refers to the filter processing by referring to the pixel values of the pixels around the processing target pixel to be filtered and the pixel values of the pixels around the boundary to be filtered. Is disclosed. However, since the filter processing method disclosed in Non-Patent Document 1 is a process in which the amount of calculation becomes very large, a technique obtained by simplifying the processing technique described in Non-Patent Document 1 is used in actual filter processing. (For example, Patent Document 1).
- the filter device calculates a value indicating the state of the block boundary before performing the filter process, as in the filter processing technique disclosed in Non-Patent Document 1. . That is, the filter device calculates ABS (p1-q0), ABS (p0-q0), and ABS (q0-q1). When the value indicating the state of the block boundary is less than a predetermined threshold, the filter device performs the filter process on the processing target pixel that is the target of the filter process. Note that if these values are equal to or greater than a predetermined threshold value, the filter processing for the processing target pixel is not executed.
- the filter device described in Patent Literature 1 uses the pixel values p1 of the pixels P1 and Q1 around the block boundary pixel. Then, linear interpolation is performed using the pixel value q1, and the pixel value p0 ′ and the pixel value q0 ′ after the filter processing are calculated for the pixel P0 and the pixel Q0 adjacent to the block boundary.
- FIG. 18 is a diagram schematically showing the filter processing in Patent Document 1.
- Patent Document 1 discloses a method of performing curvilinear interpolation by referring to more pixels.
- Non-Patent Document 2 In recent years, with the increase in definition of an encoding target image, an increase in the size of a block to be orthogonally transformed has been studied (for example, see Non-Patent Document 2). As disclosed in Non-Patent Document 2, when the size of a block to be subjected to orthogonal transformation is enlarged, a low-frequency component of an image, that is, a gradual change in the image is easily expressed.
- Non-Patent Document 1 it is necessary to refer to the pixel values in the pixels around the processing target pixel to be filtered and the pixel values of the pixels in the vicinity of the boundary to be filtered. Don't be. For this reason, when the size of a block to be subjected to orthogonal transformation is large, the amount of calculation that is enormous even with the conventional block size becomes even larger. That is, the filter processing technique described in Non-Patent Document 1 cannot be a practical means of filter processing for a block having a size larger than the conventional one.
- the filter processing method described in Patent Document 1 is an interpolation method without considering the original pixel value in the processing target pixel that is the target of the filtering process. Will not be able to interpolate properly. Therefore, the decoded image subjected to the filter process still has a problem that the image quality is deteriorated.
- the present invention has been made in view of the above-described problems, and its main purpose is to prevent deterioration in the image quality of a decoded image while suppressing an increase in the amount of calculation due to the block size. It is to provide a filter device.
- a filter device provides When decoding image information encoded by dividing into blocks of different sizes, the processing target is located near the boundary between a certain block and the neighboring block adjacent to the certain block, and belongs to the certain block.
- a filter device for correcting a pixel value of a pixel First determining means for determining whether the size of the certain block is equal to or larger than a first threshold; When the size of the certain block is equal to or larger than the first threshold value, the pixel is in the certain block included in at least one of the row and the column having the processing target pixel and is adjacent to the adjacent block Difference value calculation means for calculating a difference absolute value between a pixel value of the boundary pixel and a pixel value of an adjacent pixel adjacent to the boundary pixel in the adjacent block; A pixel that calculates a corrected pixel value of the processing target pixel based on the pixel value of the processing target pixel and the pixel value of the adjacent pixel when the calculated absolute difference value is smaller than a second threshold value And a value calculating means.
- the filter device provides a difference between a pixel value of a boundary pixel in a certain block and a pixel value of an adjacent pixel adjacent to the boundary pixel in an adjacent block.
- a pixel value calculation unit that calculates the corrected pixel value of the processing target pixel based on the pixel value before correction of the processing target pixel and the pixel value of the adjacent pixel is provided. ing.
- the corrected pixel value in the processing target pixel is calculated with reference to the pixel value of the minimum necessary pixel. That is, the filter processing can be performed on the processing target pixel with the minimum necessary amount of calculation.
- the filter device can prevent deterioration in image quality in the decoded image while suppressing an increase in the amount of calculation due to the size of the block from which block noise is to be removed. There is an effect.
- the pixel value calculation unit may further include the pixel value calculation unit when the processing target pixel is located in the vicinity of a boundary between the certain block and the adjacent block in any of the row and column directions. It is preferable to use the pixel value of the adjacent pixel adjacent to the boundary pixel in the row direction and the pixel value of the adjacent pixel adjacent to the boundary pixel in the column direction as the pixel value of the adjacent pixel.
- the horizontal filter processing and the vertical filter processing can be performed simultaneously.
- the absolute value of the difference between the pixel values adjacent to each other in the row direction is first determined from the boundary pixels adjacent to the adjacent block among the pixels included in each row of the certain block. From the boundary pixels adjacent to the adjacent block among the pixels included in each column of the certain block up to a pixel equal to or greater than the threshold value of 3, the absolute value of the pixel value difference between the pixels adjacent in the column direction is first Further, it is preferable that pixels up to the third threshold value or more are set as the processing target pixels.
- the pixel range set as the processing target pixel is appropriately set by calculating the absolute value of the difference between the pixel values adjacent in the row direction from the boundary pixel adjacent to the adjacent block. be able to.
- the filter device can select a range of pixels that can be a processing target pixel to be filtered in a row or a column, that is, a range in the vicinity of a boundary between a certain block and an adjacent block. Alternatively, there is an effect that can be appropriately set for each column.
- the filter device In the filter device according to the present invention, if the size of the certain block is smaller than the first threshold, whether the size of any adjacent block adjacent to the certain block is equal to or larger than the first threshold.
- a second determination unit that determines whether the size of any of the adjacent blocks is greater than or equal to the first threshold value. It is preferable to configure the filter strength to be higher than that in the case of being smaller than the first threshold.
- Block noise can be removed.
- the pixel value calculation means further includes the following equation (1): W1 ⁇ p + W2 ⁇ q / (W1 + W2) (1) (In the above formula (1), p is the pixel value of the pixel to be processed, q is the pixel value of the adjacent pixel, and W1 and W2 are predetermined weighting coefficients) It is preferable to calculate a pixel value after correction of the processing target pixel using.
- the pixel value calculation means further includes the following equation (2): WP ⁇ p + WH ⁇ q + WV ⁇ r / (WP + WH + WV) (2)
- p is the pixel value of the pixel to be processed
- q is the pixel value of the adjacent pixel adjacent to the boundary pixel in the row direction
- r is the pixel value of the adjacent pixel adjacent to the boundary pixel in the column direction
- WP, WH and WV indicate predetermined weighting factors
- a filter processing method When decoding image information encoded by dividing into blocks of different sizes, the processing target is located near the boundary between a certain block and the neighboring block adjacent to the certain block, and belongs to the certain block.
- a filter processing method for correcting a pixel value of a pixel A first determination step of determining whether the size of the certain block is equal to or larger than a first threshold; When the size of the certain block is equal to or larger than the first threshold value, the pixel is in the certain block included in at least one of the row and the column having the processing target pixel and is adjacent to the adjacent block.
- a difference value calculating step for calculating a difference absolute value between a pixel value of a boundary pixel that is a pixel and a pixel value of an adjacent pixel adjacent to the boundary pixel in the adjacent block; A pixel that calculates a corrected pixel value of the processing target pixel based on the pixel value of the processing target pixel and the pixel value of the adjacent pixel when the calculated absolute difference value is smaller than a second threshold value And a value calculating step.
- a decoding device and an encoding device provided with the filter device according to the present invention are also included in the category of the present invention.
- the filter device Similar effects can be achieved.
- the filter device is positioned in the vicinity of a boundary between a certain block and an adjacent block adjacent to the certain block when decoding the image information encoded by dividing into a plurality of blocks having different sizes.
- a filter device for correcting a pixel value of a processing target pixel belonging to the certain block wherein the first determining means for determining whether the size of the certain block is equal to or larger than a first threshold; When the size of a certain block is equal to or larger than the first threshold, the processing target pixel is based on the pixel value of at least one pixel belonging to the certain block and the pixel value of at least one pixel belonging to the adjacent block.
- the pixel value calculating means for calculating the corrected pixel value may be provided.
- the pixel value calculation means when the size of the certain block is equal to or larger than the first threshold, uses a smaller number of pixels than used when the size of the certain block is smaller than the first threshold. It is good also as a structure which calculates the pixel value after correction
- the filter device has the size of a certain block from which block noise is removed being equal to or larger than the first threshold value, and the pixel value of the boundary pixel in a certain block and the adjacent block
- the absolute difference between the pixel value of the adjacent pixel adjacent to the boundary pixel is smaller than the second threshold value, the pixel value before correction of the processing target pixel located near the boundary between a certain block and the boundary block, and the adjacent pixel Pixel value calculation means for calculating a corrected pixel value of the processing target pixel based on the pixel value.
- the filter device it is possible to prevent deterioration of the image quality of the decoded image while suppressing an increase in the amount of calculation due to the size of the target block having the processing target pixel. Play.
- FIG. 3 is a flowchart illustrating an operation of the filter device according to the first embodiment. It is the figure which showed typically the block division in a certain image data. It is the figure which showed typically the area
- FIG. 10 is a block diagram illustrating a main configuration of a decoding device according to Embodiment 2.
- FIG. It is the figure which showed typically the pixel in a block boundary, and the pixel value of the said pixel.
- Embodiment 1 An embodiment of a filter device according to the present invention will be described below as Embodiment 1 with reference to FIGS.
- the filter device according to the present invention is for removing block noise generated when decoding image data (image information) encoded by being divided into a plurality of blocks. Such block noise is particularly noticeable at the boundary between adjacent blocks. Therefore, the filter device according to the present invention performs a filter process for removing block noise on a processing target pixel located in the vicinity of a block boundary.
- FIG. 1 is a block diagram showing a main configuration of the filter device 1.
- the filter device 1 receives a pixel value before filter processing (a pixel value before correction) and outputs a pixel value after the filter processing (a pixel value after correction).
- the pixel value input / output unit may be a single pixel unit, or a plurality of pixel values around the block boundary in the target block having the processing target pixel to be filtered may be an input / output unit.
- the filter device 1 refers to the size of the target block having the processing target pixel to be subjected to the filtering process in order to determine the pixel value to be referred to when calculating the pixel value after the filtering process.
- the filter device 1 includes a block size determination unit 10, a difference value calculation unit 11, a pixel value calculation unit 12, an adjacent block size determination unit 13, and an output unit 14. Each member will be described below.
- the block size determination unit 10 determines whether or not the size of a target block having a processing target pixel to be subjected to filter processing is equal to or larger than a threshold Th1 in a plurality of blocks into which image data is divided.
- the threshold Th1 for example, the value of the area of the block (the total number of pixels included in the block) can be used.
- the block size determination unit 10 sets a block size of 32 ⁇ 32 among the blocks of size 8 ⁇ 8, 16 ⁇ 16, and 32 ⁇ 32 to a size equal to or larger than the threshold value Th1. It is determined that Of course, a value other than the area of the block can be used as the threshold Th1.
- the number of pixels in the vertical or horizontal direction of the block can be set as the threshold Th1.
- the “block” that is the target of the filtering process is the same as the block that is the unit of conversion or inverse conversion in the conversion or inverse conversion process. That is, the “block size” in this specification and the like is the same size as the transform size in the transform or inverse transform. That is, the block size in the filter device 1 is equal to the transform size for performing orthogonal transform in the image encoding device (not shown), and equal to the transform size for performing inverse orthogonal transform in the image decoding device (not shown). For example, when orthogonal transformation is performed using 8 ⁇ 8 DCT, the size of the block to be filtered is 8 ⁇ 8.
- the image coding apparatus performs reduction / decimation (subsampling) before orthogonal transformation, and performs enlargement / interpolation after inverse orthogonal transformation in the image decoding apparatus.
- the size of the block to be filtered is not the same size as the transform size obtained by performing the orthogonal transform as described above.
- the block size in this case is the same size as the block size before reduction / decimation before orthogonal transformation (block size after enlargement / interpolation). Note that the block size is not a scalar value but a vector value represented by horizontal x vertical.
- a ⁇ A representing the block size indicates the number of pixels in the horizontal direction and the number of pixels in the vertical direction in the block. That is, “8 ⁇ 8” in this specification and the like means a block composed of a total of 64 pixels of 8 pixels in the horizontal direction and 8 pixels in the vertical direction.
- the difference value calculation unit 11 calculates a difference absolute value of pixel values. More specifically, the difference absolute value between the pixel value of the boundary pixel that is a pixel adjacent to the pixel of the adjacent block adjacent to the target block and the pixel value of the adjacent pixel adjacent to the boundary pixel in the adjacent block is calculated. In some cases, an absolute difference value of pixel values of pixels in the target block is also calculated. The difference absolute value calculated by the difference value calculation unit 11 is determined according to the size of the target block. What kind of absolute value is calculated in which case will be described in detail below.
- adjacent in this specification and the like means contacting with a reference pixel or block in the vertical direction (vertical direction) or the horizontal direction (horizontal direction). That is, a block or pixel that is in contact with a reference pixel or block in an oblique direction is not included in the category of adjacent pixels or blocks in this specification and the like.
- difference absolute value in this specification and the like means an absolute value of a difference between pixel values.
- the pixel value calculation unit 12 calculates a post-processing pixel value after the filtering process on the processing target pixel by filtering the processing target pixel in the target block.
- the calculation method for calculating the post-processing pixel value from the pre-processing pixel value before the filter processing of the processing target pixel is determined according to the size of the target block. Which calculation method is used in which case will be described in detail below.
- the adjacent block size determination unit 13 determines whether the size of any adjacent block adjacent to the target block is greater than or equal to the threshold Th1 when the size of the target block is smaller than the threshold Th1. That is, it is determined whether there is a block having a size equal to or larger than the threshold Th1 in a block adjacent to the target block.
- the output unit 14 replaces the pre-processing pixel value of the processing target pixel with the post-processing pixel value calculated by the pixel value calculation unit 12. That is, the pixel value replacement unit replaces the pixel value of the processing target pixel.
- the output destination of the processed pixel value is appropriately changed according to the device provided with the filter device 1 and the application of the filter device 1. For example, when the filter device 1 is provided as a loop filter in the image decoding device, the output unit 14 outputs the frame image in which the decoded image is stored.
- FIG. 2 is a flowchart showing an outline of the operation of the filter device 1.
- the filter device 1 sets a target block having processing target pixels to be subjected to filter processing (step S1). Subsequently, the block size determination unit 10 determines whether or not the size of the target block is greater than or equal to the threshold Th1 (step S2). When the size of the target block is equal to or larger than the threshold Th1 (Yes in Step S2), the difference value calculation unit 11 calculates a difference absolute value (Step S3). The absolute difference value calculated in step S3 will be described in detail below.
- the pixel value calculation unit 12 calculates the post-processing pixel value of the processing target pixel when the difference absolute value calculated in step S3 is smaller than the threshold Th2 (second threshold) (step S4). If the absolute difference value is equal to or greater than the threshold Th2, the pixel value calculation unit 12 does not calculate the post-processing pixel value of the processing target pixel. That is, the filter device 1 stops the filter process on the processing target pixel. Finally, the output unit 14 outputs the calculated post-processing pixel value (step S5).
- the processing of the filter device 1 when the size of the target block is equal to or larger than the threshold Th1 is hereinafter referred to as a pattern A.
- the adjacent block size determination unit 13 determines whether the size of the adjacent block adjacent to the target block is equal to or larger than the threshold Th1 (Step S6).
- the difference value calculation unit 11 calculates a difference absolute value (Step S7). At this time, the difference absolute value calculated by the difference value calculation unit 11 is different from the difference absolute value calculated in step S3. The absolute difference value calculated in step S7 will be described in detail below.
- the pixel value calculation unit 12 calculates a post-processing pixel value of the processing target pixel when the difference absolute value calculated in step S7 is smaller than a predetermined threshold (step S8). Note that the method for calculating the processed pixel value in step S8 is different from the method for calculating the processed pixel value in step S4. In addition, when the difference absolute value is greater than or equal to a predetermined value, the pixel value calculation unit 12 does not calculate the post-processing pixel value of the processing target pixel. That is, the filter process for the processing target pixel is stopped.
- the output unit 14 outputs the calculated post-processing pixel value (step S5).
- the processing of the filter device 1 when the size of the target block is less than the threshold Th1 and the size of the adjacent block is greater than or equal to the threshold Th1 is hereinafter referred to as a pattern B.
- the difference value calculation unit 11 calculates the absolute difference value (step S9). At this time, the difference absolute value calculated by the difference value calculation unit 11 is different from the difference absolute value calculated in step S3 and step S7. The difference absolute value calculated in step S9 will be described in detail below.
- the pixel value calculation unit 12 calculates a post-processing pixel value of the processing target pixel when the difference absolute value calculated in step S9 is smaller than a predetermined threshold (step S10). Note that the method for calculating the processed pixel value in step S10 is different from the method for calculating the processed pixel value in steps S4 and S8. In addition, when the difference absolute value is greater than or equal to a predetermined value, the pixel value calculation unit 12 does not calculate the post-processing pixel value of the processing target pixel. That is, the filter process for the processing target pixel is stopped.
- the output unit 14 outputs the calculated post-processing pixel value (step S5).
- the processing of the filter device 1 when the size of the target block is less than the threshold value Th1 and the size of the adjacent block is also less than the threshold value Th1 is hereinafter referred to as a pattern C.
- the processing of the filter device 1 is divided into three processing patterns (patterns A to C) according to the block size of the target block and the size of the adjacent block size adjacent to the target block size. Details of these three processes will be described below.
- Pattern A When the size of the target block is greater than or equal to the threshold Th1
- the details of the calculation method (pattern A) of the pixel value of the processing target pixel when the size of the target block having the processing target pixel of the filter processing is equal to or larger than the threshold Th1 will be described below with reference to FIGS.
- FIG. 3 is a diagram schematically illustrating an example of block division in an image.
- the image to be filtered is divided into three size blocks, a 32 ⁇ 32 size block, a 16 ⁇ 16 size block, and an 8 ⁇ 8 size block. Yes. That is, the image data shown in FIG. 3 is divided by blocks of different sizes.
- FIG. 3 is a diagram schematically showing a region having pixels to be filtered in a block whose target block size is equal to or larger than the threshold Th1.
- FIG. 5 is a diagram schematically showing pixels included in a region indicated by hatching in FIG. That is, it is a diagram schematically showing pixels at a boundary portion of a block.
- the pixel Pi of one block is the pixel Pi
- the pixel of the other block is the pixel Qi (i is 0 to n, n is an arbitrary integer). Note that the value of i in the pixel Pi and the pixel Qi becomes smaller as it is closer to the block boundary.
- pixel values corresponding to the pixel Pi and the pixel Qi are shown as a pixel value pi and a pixel value qi, respectively, and the processed pixel value corresponding to the unprocessed pixel value pk in the processing target pixel Pk. It is shown as pk ′.
- k is an integer equal to or less than the threshold Th3.
- a detailed method for setting the threshold Th3 will be described in detail below, and the description thereof is omitted here.
- the difference value calculation unit 11 calculates the absolute difference value between the pixel value of the boundary pixel and the pixel value of the adjacent pixel, and the absolute difference value calculated by the pixel value calculation unit 12 is smaller than the threshold Th2.
- the corrected pixel value of the processing target pixel is calculated based on the pixel value of the processing target pixel and the pixel value of the adjacent pixel.
- the difference value calculation unit 11 has a pixel value p0 of the boundary pixel P0 that is a pixel of the target block adjacent to a pixel of the adjacent block adjacent to the target block, and a boundary in the adjacent block.
- the pixel value calculation unit 12 is based on the pixel value pk before processing of the processing target pixel Pk and the pixel value q0 of the adjacent pixel Q0.
- a pixel value pk ′ after processing of the processing target pixel Pk is calculated.
- W1 and W2 are predetermined filter coefficients (weighting coefficients) set according to the number of pixels from the boundary pixels, and are values satisfying 0 ⁇ W1 and W2 ⁇ W1 + W2.
- FIG. 6 is a diagram illustrating an example of filter coefficients in the pixels P0 to P7 and the pixel Q0.
- the output unit 14 outputs the processed pixel value pk ′ calculated by the pixel value calculation unit 12. Note that the destination to which the output unit 14 outputs the filtered pixel value pk ′ is appropriately changed according to the type of device on which the filter device 1 is mounted.
- the processing target pixel is the pixel Pk
- the processing target pixel may be the pixel Qk as a matter of course.
- the pixel value pk and the pixel value q0 in the above formula (2) may be read as the pixel value qk and the pixel value p0, respectively.
- the pixel value pk of the processing target pixel Pk and the pixel value q0 of the adjacent pixel Q0 are used only after the processing of the processing target pixel Pk.
- a pixel value pk ′ is calculated. This is because, in a block having a large block size, even if the pixel value pk ′ after processing is calculated using the pixel values of pixels around the processing target pixel Pk, the calculated pixel value pk ′ hardly changes. Because.
- an area including a high frequency component (an area where the change in pixel value is large) is encoded with a small size block
- An area including a frequency component is encoded with a large-size block.
- a block that greatly exceeds the size of 8 ⁇ 8 such as a block of 32 ⁇ 32 size, often has a region composed of only low frequency components.
- the difference between the pixel value pk of the processing target pixel Pk and the pixels around the processing target pixel Pk (for example, the pixel value pk-1 of the pixel Pk-1 and the pixel value pk + 1 of the pixel Pk + 1). Becomes smaller. Therefore, when the pixel value pk ′ after processing is calculated, the calculated pixel value hardly changes even if the pixels around the processing target pixel Pk are referred to in addition to the pixel value pk of the processing target pixel Pk. .
- a pixel value around the pixel is used for filter processing of a certain pixel, it is possible to suppress changes in the pixel value.
- this effect is effective for pixels near the block boundary, but is not necessary for pixels far from the block boundary.
- the number of processing target pixels that are far from the block boundary is larger than that of a small-sized block, so that the number of processing target pixels that do not need to refer to pixel values of pixels around the processing target image increases.
- the filter device 1 may calculate the pixel value pk ′ after processing of the processing target pixel Pk from only the pixel values of the two pixels, the pixel value pk of the processing target pixel Pk and the pixel value q0 of the adjacent pixel Q0. .
- the filter device 1 can prevent deterioration of the image quality in the decoded image while suppressing an increase in the amount of calculation due to the size of the block.
- prevention in this specification etc. It does not mean that the deterioration of image quality is completely prevented, but it means that the image quality is suppressed even a little compared with the conventional case.
- FIG. 7 is a diagram schematically showing a region having pixels to be subjected to filter processing in a block in which the size of the target block is less than the threshold Th1 and the size of an adjacent block adjacent to the target block is equal to or larger than the threshold Th1.
- the pixel Qk shown in FIG. 5 is a processing target pixel
- the pixel value after processing corresponding to the pixel value qk before processing becomes the pixel value qk ′.
- k is an integer equal to or less than the threshold Th3.
- the difference value calculation unit 11 is adjacent to the pixel value q0 of the boundary pixel Q0 that is a pixel of the target block adjacent to the pixel of the adjacent block adjacent to the target block, and adjacent to the boundary pixel Q0 in the adjacent block.
- the pixel value calculation unit 12 determines whether or not the difference absolute value d and the difference absolute value ap calculated by the difference value calculation unit 11 are less than predetermined threshold values ⁇ 1 and ⁇ 1, respectively. That is, the pixel value calculation unit 12 determines whether d ⁇ 1 and ap ⁇ 1.
- the pixel value calculation unit 12 processes the pixel value p0 of the adjacent pixel P0 and the processing target pixel Qk. Based on the previous pixel value qk and the pixel values of the pixels around the processing target pixel Qk, the processed pixel value qk ′ of the processing target pixel Qk is calculated.
- the output unit 14 outputs the processed pixel value qk ′ calculated by the pixel value calculation unit 12.
- the processing of the pattern C may be performed by the same method as the processing in the conventional filter device.
- the processing of pattern C will be specifically described below.
- the case where the pixel Pk shown in FIG. 5 is the processing target pixel Pk will be described as an example. That is, the processed pixel value corresponding to the unprocessed pixel value pk is pk ′.
- k is an integer equal to or less than the threshold Th3.
- the difference value calculation unit 11 has a pixel value p0 of the boundary pixel P0 that is a pixel of the target block adjacent to the pixel of the adjacent block adjacent to the target block, and a pixel value of the adjacent pixel Q0 that is adjacent to the boundary pixel P0 in the adjacent block.
- the pixel value calculation unit 12 determines whether the difference absolute value d and the difference absolute value ap calculated by the difference value calculation unit 11 are less than predetermined threshold values ⁇ 2 and ⁇ 2, respectively. That is, the pixel value calculation unit 12 determines whether d ⁇ 2 and ap ⁇ 2.
- the pixel value calculation unit 12 determines the pixel value pk before the processing of the processing target pixel Pk and the processing target. Based on the pixel values of the pixels around the pixel Pk, the pixel value pk ′ after processing of the processing target pixel Pk is calculated.
- the output unit 14 outputs the processed pixel value pk ′ calculated by the pixel value calculation unit 12.
- the processing target pixel may be the pixel Qk.
- the block noise at the block boundary may be more significant than when the size of the adjacent block is less than the threshold Th1.
- the change in the pixel value in the block is small, that is, the image is flat. Therefore, in the filter process in the pattern B, it is preferable to make the filter strength stronger than that in the pattern C.
- the threshold values ⁇ 1 and ⁇ 1 in the pattern B are compared with the threshold values ⁇ 2 and ⁇ 2 in the pattern C, it is preferable that ⁇ 1> ⁇ 2 and ⁇ 1> ⁇ 2. As a result, the pattern B process is more likely to be filtered, so that a stronger filter can be applied than the pattern C process.
- the value of the filter coefficient multiplied by each pixel value in the pattern B process may be set so that the low-pass effect is higher than that in the pattern C process. Even in this case, compared to the pattern C process, the process of the pattern B can perform a process with a stronger filter strength on the processing target pixel.
- the adjacent block adjacent to the target block is adjacent even when the change in the pixel value of the component perpendicular to the boundary direction is large, that is, when the change around the pixel value q0 is large. Only the pixel value q0 of the pixel Q0 is used.
- the pixel value of the adjacent pixel in the adjacent block adjacent to the target block when the change around the pixel value is large, the pixel value may be changed by the filtering process. In this case, the target block to be filtered may be undulated (noise on the line is generated). Such noise is particularly noticeable when the filter target range is wide, and the decoded image is likely to appear visually unnatural.
- the pixel value q0 of the adjacent pixel Q0 is not used as it is for the calculation of the pixel value pk ′ of the processing target pixel Pk, but the pixel value of the adjacent pixel that has been filtered in advance is used. Is preferred.
- a filter process (a modified example of the process of pattern A) when the adjacent pixels are previously filtered will be specifically described below.
- FIG. 8 is an expanded view of FIG. 5 and shows not only one boundary adjacent to the target block but also the entire boundary adjacent to the target block.
- a pixel of one block is a pixel Pij
- a pixel of the other block is a pixel Qij (i is 0 to n, n is an arbitrary integer). Note that the value of i in the pixel Pij and the pixel Qij becomes smaller as it is closer to the block boundary.
- j is an index for expressing a component perpendicular to the boundary, and takes a value from 0 to [the number of pixels in the vertical direction of the target block ⁇ 1].
- the pixel value q0j of each adjacent pixel Q0j of the adjacent block is filtered in advance, and the pixel value q0j ′ after the filtering process is referred to.
- the pixel value q0j ′ of the pixel Q0j may be calculated so that a low-pass effect is applied between pixels perpendicular to the boundary direction.
- q0k ′ (5 ⁇ q0k ⁇ 1 + 6 ⁇ q0k + 5 ⁇ q0k + 1) / 16 may be used.
- k is an arbitrary integer within the range of j.
- q0k ′′ (5 ⁇ q0k ⁇ 1 ′ + 6 ⁇ q0k ′ + 5 ⁇ q0k + 1 ′) / 16, which is a calculation formula using the pixel value q0k ′ of the pixel Q0k. It is preferable to use the pixel value q0k ′′ calculated by the above in place of the pixel value q0k ′.
- the adjacent pixel values q0j ′, q0j ′′, etc. of the adjacent block are calculated in advance before the filtering process.
- the threshold value Th3 is a pixel included in a row or column in a certain block, and a difference absolute value of a pixel value between pixels adjacent in the row or column direction is first calculated from a boundary pixel adjacent to an adjacent block adjacent to a certain block. This is the number of pixels up to pixels that are equal to or greater than the threshold Th4 (third threshold).
- the threshold value Th3 may be a preset value or a value set by a predetermined algorithm.
- the filter device 1 first calculates the absolute value of the difference between the pixel values adjacent to each other in the row direction or the column direction from the boundary pixel adjacent to the adjacent block among the pixels included in each row or each column of a certain block.
- the pixels to be processed are pixels up to the threshold Th4 or more.
- the “number of pixels” in this specification and the like means a value indicating the number of pixels from the boundary pixel as a base point. Specifically, referring to FIG. 5, the pixel having two pixels becomes the pixel P2 when the pixel P0 is the base point. Note that the pixel having 0 pixels is the pixel P0.
- FIG. 9 is a flowchart showing a method for setting the threshold Th3.
- the pixel Pk in this section indicates an arbitrary pixel in the target block, and the pixel value pk indicates the pixel value of the pixel Pk.
- K is an arbitrary integer.
- the pixel of k 0, that is, the pixel P0 indicates the boundary pixel in the target block, and the distance from the boundary pixel increases as the value of k increases. That is, the value of k matches the value of the number of pixels in the pixel Pk.
- the filter device 1 sets the value of k to “0”, that is, sets the target pixel as the pixel P0 (step S20). Subsequently, the difference value calculation unit 11 calculates a difference absolute value between the pixel value p0 of the pixel P0 and the pixel value q0 of the adjacent pixel Q0 adjacent to the pixel P0 in the adjacent block adjacent to the target block (step S21). . And the difference value calculation part 11 determines whether the calculated difference absolute value is more than threshold value Th2 (step S22). When the calculated difference absolute value is greater than or equal to the threshold Th2 (Yes in step S22), the filter device 1 stops the filter process for the processing target pixel. This is the same as the process in step S4 described in the operation of the filter device 1 (FIG. 2).
- the filter device 1 When the calculated difference absolute value is less than the threshold Th2 (No in step S22), the filter device 1 adds “1” to the value of k (step S23). Subsequently, the difference value calculation unit 11 calculates an absolute difference value between the pixel value pk of the pixel Pk and the pixel value p (k ⁇ 1) of the pixel P (k ⁇ 1) (step S24). Next, the filter device 1 determines whether or not the difference absolute value calculated in step S24 is greater than or equal to the threshold Th4 (step S25).
- the threshold value Th4 is a value set in advance in the filter device 1. The threshold value Th4 may be a value fixed in the filter device 1 or a value set as appropriate by the user.
- step S27 If the value of k at that time is less than the predetermined value Tn (Yes in step S27), the filter device 1 adds “1” to the value of k at that time (step S23), and repeats the processing from step S24. Execute. In this way, the filter device 1 repeats the process until the absolute difference value calculated in step S24 is equal to or greater than the threshold value Th4 or the value of k is equal to or greater than the predetermined value Tn.
- the predetermined value Tn is a value set in advance in the filter device 1.
- the predetermined value Tn sets a limit value of pixels that can be processed pixels in the target block.
- the difference absolute value calculated in step S24 is less than the threshold Th4.
- the value is set in order to prevent the pixel that does not need to be filtered from being set to be filtered.
- a prediction image close to an encoding target image is generated, and a prediction error that is a difference between the encoding target image and the prediction image is converted and encoded. Therefore, the pixel value of the target block that is the target of the filtering process is decoded by the sum of the predicted image and the prediction error (the signal after inverse transformation). If the block size when executing the conversion is large, the change in the pixel value in the prediction error block can be expected to be small, but if there is a large change in the prediction image, the sum of the prediction image and the prediction error is expected.
- the change in the pixel value in the decoded image obtained by the above is not necessarily small. That is, when the change of the pixel value in the decoded image is large due to the influence of the predicted image, it is necessary to set the range to be subjected to the filtering process to be small.
- Non-Patent Document 1 and Patent Document 1 it is possible to determine whether or not to perform uniform filter processing on the block boundary, but filter processing is performed on a part of the block boundary. It cannot be processed.
- the configuration of the filter device 1 according to the present embodiment can solve the above problems. That is, as described above, the filter device 1 can appropriately set the value of the threshold Th3 in the row or column of the target block by calculating the difference absolute value between adjacent pixels. Therefore, when it is considered that there is a change in the pixel value in the decoded image, it is possible to appropriately change the pixel range to be filtered in the row or column of the target block. Of course, when there is no change in the decoded image, up to a predetermined number of pixels may be filtered.
- Non-Patent Document 1 and Patent Document 1 values corresponding to conversion of pixel values between pixels (for example, ABS (p2-p0), ABS ( p1-p0)) is calculated, and using this value, it is determined whether or not to perform the filter processing. However, this determination only determines whether or not to perform filtering on the entire boundary to be processed, and does not determine how many pixels away from the block boundary are to be filtered. Absent.
- Modification of threshold value Th3 setting method Next, a modified example of the method for setting the threshold Th3 will be described.
- This modification is a method that takes into account the case where the change in the pixel value of the component perpendicular to the boundary direction in the adjacent block adjacent to the target block is large, that is, the case where the change in the pixel value around the pixel Q0 is large.
- the threshold value Th3 and the predetermined value Tn when filtering the pixel value pij of the pixel Pij in FIG. 8 are set as the threshold value Th3j and the predetermined value Tnj, respectively.
- the magnitude MAGj of change around the pixel value q0j is calculated based on the sum of absolute differences of values around the pixel value q0j.
- MAGk ABS (q0k ⁇ q0k ⁇ 1) + ABS (q0k ⁇ q0k + 1) Calculate using.
- k is an arbitrary integer within the range of j.
- MAGk ′ (MAGk ⁇ 1 + 2 ⁇ MAGk + 2 ⁇ MAGk + 1) / 4 It is preferable to use MAGk ′ calculated by using MAGk instead of MAGk.
- a value obtained by applying a low-pass filter to the threshold Th3j may be used instead of the threshold Th3j.
- Th3k Threshold Th3k ′ (Th3k ⁇ 1 + 2 ⁇ Th3k + 2 ⁇ Th3k + 1) / 4
- the threshold value Th3k ′ calculated by using may be used.
- the predetermined threshold values Th2, Th4, Tn, ⁇ , ⁇ , ⁇ 1, ⁇ 1, ⁇ 2, ⁇ 2, ⁇ 1, and ⁇ 2 used in the above description are preferably changed according to the quantization parameter. That is, it is preferable that the threshold value be changed so as to increase as the quantization parameter increases.
- the range to be processed pixels in the target block is changed according to the size of the block. That is, if the threshold value Th3 is not set due to the boundary state, the threshold value Th3 may be determined for each block size. In this case, the threshold value Th3 is set to increase as the block size increases. In this case, the predetermined value Tn may be determined for each block size so that the predetermined value Tn increases as the block size increases.
- the filter processing techniques described in Non-Patent Document 1 and Patent Document 1 sequentially process two block boundaries (vertical boundary and horizontal boundary) in a target block having a processing target pixel to be filtered. To do.
- the filter processing techniques described in Non-Patent Document 1 and Patent Document 1 in the filter processing technique that sequentially filters the vertical boundary and the horizontal boundary, the vertical boundary and the horizontal boundary are simultaneously set. Compared with the filter processing, the image quality of the decoded image is deteriorated.
- horizontal filtering and vertical filtering are performed. Will be applied twice.
- the smoothing effect by the filter process becomes too strong, and the image quality of the decoded image is deteriorated.
- the weight between the horizontal direction and the vertical direction is different, the image quality of the decoded image is deteriorated.
- the filtering process in the horizontal direction and the filtering process in the vertical direction at the same time in a region close to both the horizontal and vertical boundaries of the adjacent blocks such as the corners of the blocks. That is, the filtering process for the horizontal direction and the filtering process for the vertical direction are simultaneously performed in a single process, thereby preventing the filtering process from being performed twice.
- the filtering process performed simultaneously in the horizontal direction and the vertical direction is referred to as “planar filtering process”.
- FIG. 10 is a diagram showing an area close to both the horizontal and vertical boundaries of adjacent blocks in the target block. More specifically, the pixels included in the area indicated by the hatched portion in the block X (target block) are below the threshold Th3 from the boundary pixel at the boundary between the block X and the block Y (horizontal boundary), In addition, the pixel is equal to or less than the threshold Th3 from the boundary pixel at the boundary (vertical boundary) between the block X and the block Z.
- FIG. 11 shows an example of the pixel in the region indicated by the oblique lines in FIG.
- a pixel Pij shown in FIG. 11 is a pixel included in the area indicated by the oblique line of the block X in FIG.
- an adjacent pixel located closest to the pixel Pij in the row direction is indicated as a pixel Q0j
- an adjacent pixel located closest to the pixel Pij in the column direction is indicated as a pixel Ri0.
- the adjacent block having the pixel Q0j is a block Z adjacent to the block X in the horizontal direction
- the adjacent block having the pixel Ri0 is a block Y adjacent to the block X in the vertical direction
- the pixel Pij in the block X is a pixel that is i pixels away from the pixel Q0j in the row direction and j pixels away from the pixel Ri0 in the column direction.
- i and j are both values below the threshold Th3.
- “row” means a vertical line in the block
- “column” means a horizontal line in the block.
- the number of pixels from the boundary pixel at the boundary in the horizontal direction is the number of pixels from the boundary pixel at the boundary in the vertical direction as shown in FIG. Is preferably a pixel having a threshold value Th3 or less. Therefore, when the filtering process is performed in the area shown in FIG. 4, the planar filtering process is only the area shown by the oblique lines in FIG. In other words, horizontal or vertical filter processing is performed on the area other than the area indicated by the oblique lines in FIG. 12 and the area indicated by the oblique lines in FIG.
- the processing target pixel Pij is included in a block of 32 ⁇ 32 size
- the adjacent pixel in the row direction of the pixel Pij is the pixel Q0j
- the adjacent pixel in the column direction of the pixel Pij is the pixel Ri0.
- a case will be described as an example.
- the pixel value of the pixel Pij before the filtering process is the pixel value pij
- the pixel value of the pixel Q0j is the pixel value q0j
- the pixel value of the pixel Ri0 is ri0
- the pixel value of the pixel Pij after the filtering process is the pixel value.
- the method for calculating the pixel value after processing of the processing target pixel in the planar filter processing is basically the same as the method for calculating the pixel value after processing in the pattern A.
- the pixel value of the adjacent pixel adjacent to the boundary pixel in the row direction and the pixel value of the adjacent pixel adjacent to the boundary pixel in the column direction are used as the pixel value of the adjacent pixel. That is, the pixel values of two adjacent pixels, the pixel value q0j of the adjacent pixel Q0j and the pixel value ri0 of the adjacent pixel Ri0 are used.
- the pixel value calculation unit 12 calculates the pixel value pij ′ of the processing target pixel Pij based on the pixel value pij ′, the pixel value q0j, and the pixel value ri0. More specifically, the pixel value calculation unit 12 uses the following formula (7).
- pij ′ (WP ⁇ pij + WV ⁇ q0j + WH ⁇ ri0) / (WP + WV + WH) (7) Is used to calculate the processed pixel value pij ′.
- WP, WV, and WH are filter coefficients
- WH is a filter coefficient set according to the distance (number of pixels) from the boundary pixel P0j in the row direction
- WV is from the boundary pixel Pi0 in the column direction.
- the filter coefficient is set according to the distance (number of pixels)
- WP 32 ⁇ (WH or WV).
- the value of WH or WV used in the calculation of WP is the value with the larger number of pixels. That is, when the processing target pixel has 3 pixels in the horizontal direction and 1 pixel in the vertical direction, the value of WH is used. Examples of WP, WV, and WH are shown in FIGS.
- Such filter coefficients WP ′, WV ′, and WH ′ are expressed by the following equations (8) to (11).
- WV ′ m ⁇ WV (8)
- WH ′ m ⁇ WH (9)
- WP ' 1024- (WV' + WH ') (10)
- m 1024 / (WP + WV + WH) (11)
- m may be set with reference to a table registered in association with the value of (WP + WV + WH) and the value of m.
- pij ′ (WP ′ ⁇ pij + WV ′ ⁇ q0j + WH ′ ⁇ ri0) >> 10 (12)
- the pixel value pij ′ is calculated so that
- FIG. 14 and (b) in FIG. 14 are diagrams showing examples of filter coefficients WV ′ and WP ′, respectively.
- WV ′ filter coefficients
- WP ′ filter coefficients
- the filter device 1 prevents the two-dimensional filter process and the vertical filter process from being performed twice on the processing target pixel by performing the planar filter process. can do.
- FIG. 15 is a block diagram illustrating a main configuration of the image decoding device 100 according to the second embodiment.
- symbol is attached
- the image decoding device 100 includes a variable length coding / decoding unit 101, an inverse quantization / inverse transform unit 102, a predicted image generation unit 103, a frame memory 104, a block size management unit 105, and a filter device 1. I have. Members other than the filter device 1 in the image decoding device 100 can use the same members as those in the conventional image decoding device.
- the variable length encoding / decoding unit 101 performs variable length decoding processing on encoded data representing an encoding parameter such as prediction residual data of a decoding target block.
- the inverse quantization / inverse transform unit 102 inversely quantizes and inversely transforms the input parameter. That is, the inverse quantization / inverse transform unit 102 inversely quantizes and outputs the quantized coefficient of the decoded prediction residual data, and outputs the transform coefficient of the prediction residual data by inverse orthogonal transformation.
- the block size management unit 105 stores the transform size decoded by the variable length decoding process.
- orthogonal transformation is common as a transformation method in inverse transformation, it is not limited to this. For example, a method using a filter bank typified by QMF may be used, transformation using a non-orthogonal basis typified by a Gabor function, or wavelet transformation may be used.
- the signal after inverse transformation that has been inversely transformed by the inverse quantization / inverse transformation unit 102 is added to the predicted image generated by the predicted image generation unit 103, and is output as a decoded image.
- the output decoded image is stored in the frame memory 104.
- the filter device 1 acquires the block size of the target block having the processing target pixel and the block size of the block adjacent to the target block from the block size management unit 105. Further, the filter device 1 refers to the pixel value recorded in the frame memory 104 and outputs the pixel value after the filtering process. That is, the pixel value of the pixel filtered by the filter device 1 is recorded in the frame memory 104. At this time, in the frame memory 104, the pixel value after the filter process in the filter device 1 is replaced with the pixel value before the process.
- the filtered image stored in the frame memory 104 is used by the predicted image generation unit 103 to generate a predicted image.
- the image filtered by the filter device 1 is output as an output image outside the image decoding device 100. That is, the filter device 1 can be used as a loop filter (that is, a filter related to an image stored in a frame memory) in the image decoding device 100, or can be used as a post filter (that is, a filter related to an output image). You can also.
- the filter device 1 may be provided as a loop filter provided in the image encoding device.
- members other than the filter device 1 can be the same members as those of a conventionally known image encoding device.
- Each block included in the filter device 1 and the image decoding device 100 may be configured by hardware logic. Alternatively, it may be realized by software using a CPU (Central Processing Unit) as follows.
- CPU Central Processing Unit
- the filter device 1 and the image decoding device 100 include a CPU that executes instructions of a program that realizes each function, a ROM (Read Only Memory) that stores the program, and a RAM (Random Access that expands the program into an executable format. Memory) and a storage device (recording medium) such as a memory for storing the program and various data.
- a predetermined recording medium such as a hard disk drive, a solid state drive, or a hard disk drive, or a hard disk drive, or a hard disk drive, or a hard disk drive, and a hard disk drive, or a hard disk drive, or a hard disk drive, or a hard disk drive, or a hard disk drive, or a hard disk drive, or a hard disk drive, or a hard disk drive, or a hard disk drive, or a hard disk drive, or a hard disk drive, or a hard disk drive, a hard disk drive, a hard disk drive, a hard disk drive, or a hard disk drive, or a hard disk
- this recording medium records the program codes (execution format program, intermediate code program, source program) of the programs of the filter device 1 and the image decoding device 100, which are software that realizes the above-described functions, such that they can be read by a computer. Good.
- This recording medium is supplied to the filter device 1 and the image decoding device 100.
- the filter device 1 and the image decoding device 100 or CPU or MPU as a computer may read and execute the program code recorded on the supplied recording medium.
- the recording medium that supplies the program code to the filter device 1 and the image decoding device 100 is not limited to a specific structure or type. That is, the recording medium includes, for example, a tape system such as a magnetic tape and a cassette tape, a magnetic disk such as a floppy (registered trademark) disk / hard disk, and an optical disk such as a CD-ROM / MO / MD / DVD / CD-R. System, a card system such as an IC card (including a memory card) / optical card, or a semiconductor memory system such as a mask ROM / EPROM / EEPROM / flash ROM.
- a tape system such as a magnetic tape and a cassette tape
- a magnetic disk such as a floppy (registered trademark) disk / hard disk
- an optical disk such as a CD-ROM / MO / MD / DVD / CD-R.
- a card system such as an IC card (including a memory card) / optical card, or a
- the object of the present invention can be achieved even if the filter device 1 and the image decoding device 100 are configured to be connectable to a communication network.
- the program code is supplied to the filter device 1 and the image decoding device 100 via the communication network.
- the communication network is not limited to a specific type or form as long as it can supply program codes to the filter device 1 and the image decoding device 100.
- the Internet intranet, extranet, LAN, ISDN, VAN, CATV communication network, virtual private network, telephone line network, mobile communication network, satellite communication network, etc. may be used.
- the transmission medium constituting the communication network may be any medium that can transmit the program code, and is not limited to a specific configuration or type.
- a specific configuration or type for example, even wired such as IEEE 1394, USB (Universal Serial Bus), power line carrier, cable TV line, telephone line, ADSL (Asymmetric Digital Subscriber Line) line, infrared such as IrDA or remote control, Bluetooth (registered trademark), 802.11
- radio such as radio, HDR, mobile phone network, satellite line, terrestrial digital network.
- the present invention can also be realized in the form of a computer data signal embedded in a carrier wave in which the program code is embodied by electronic transmission.
- the filter device according to the present invention can be suitably used as a filter device that operates as a loop filter or a post filter in an image or video encoding device and an image or video decoding device.
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
- Image Processing (AREA)
Abstract
Description
d=ABS(p0-q0)・・・(13)
ap=ABS(p2-q0)・・・(14)
を用いて境界の状態を示す値「d」および「ap」を算出する。
p0´=(p2+2×p1+2×p0+2×q0+q1)/8
p1´=(p2+p1+p0+q0)/4
p2´=(2×p3+3×p2+p1+p0+q0)/8
q0´=(q2+2×q1+2×q0+2×p0+p1)/8
q1´=(q2+q1+q0+p0)/4
q2´=(2×q3+3×q2+q1+q0+p0)/8
を用いて、フィルタ処理後の画素値を算出する。なお、下記式中の画素値p0´、p1´、p2´、q0´、q1´およびq2´は、フィルタ処理後の各処理対象画素における画素値を示している。また、図17の(a)および図17の(b)は、フィルタ処理を施す際の各画素値におけるフィルタ係数を示した図である。
p0´=p1+(p1-q1)/3・・・(15)
q0´=q0+(q1-p1)/3・・・(16)
を用いてフィルタ処理後の画素P0´の画素値p0´および画素Q0´の画素値q0´を算出する。
サイズの異なる複数のブロックに分割して符号化された画像情報を復号する際に、あるブロックと当該あるブロックに隣接する隣接ブロックとの境界近傍に位置し、かつ、上記あるブロックに属する処理対象画素の画素値を補正するフィルタ装置であって、
上記あるブロックのサイズが第1の閾値以上であるか否かを判定する第1の判定手段と、
上記あるブロックのサイズが上記第1の閾値以上である場合、上記処理対象画素を有する行および列の少なくともいずれかに含まれる上記あるブロック内の画素であり、かつ、上記隣接ブロックに隣接する画素である境界画素の画素値と、上記隣接ブロックにおいて当該境界画素に隣接する隣接画素の画素値との差分絶対値を算出する差分値算出手段と、
算出した上記差分絶対値が第2の閾値よりも小さい場合、上記処理対象画素の画素値と、上記隣接画素の画素値とに基づいて、当該処理対象画素の補正後の画素値を算出する画素値算出手段と、を備えていることを特徴としている。
W1×p+W2×q/(W1+W2)・・・(1)
(上記式(1)中、pは上記処理対象画素の画素値、qは上記隣接画素の画素値、W1およびW2は所定の重み付け係数を示す)
を用いて上記処理対象画素の補正後の画素値を算出することが好ましい。
WP×p+WH×q+WV×r/(WP+WH+WV)・・・(2)
(上記式(2)中、pは上記処理対象画素の画素値、qは行方向の境界画素に隣接する隣接画素の画素値、rは列方向の境界画素に隣接する隣接画素の画素値、WP、WHおよびWVは所定の重み付け係数を示す)
を用いて上記処理対象画素の補正後の画素値を算出することが好ましい。
サイズの異なる複数のブロックに分割して符号化された画像情報を復号する際に、あるブロックと当該あるブロックに隣接する隣接ブロックとの境界近傍に位置し、かつ、上記あるブロックに属する処理対象画素の画素値を補正するフィルタ処理方法であって、
上記あるブロックのサイズが第1の閾値以上であるか否かを判定する第1の判定ステップと、
上記あるブロックのサイズが上記第1の閾値以上である場合、上記処理対象画素を有する行および列の少なくともいずれかに含まれる上記あるブロック内の画素であって、かつ、上記隣接ブロックに隣接する画素である境界画素の画素値と、上記隣接ブロックにおいて当該境界画素に隣接する隣接画素の画素値との差分絶対値を算出する差分値算出ステップと、
算出した上記差分絶対値が第2の閾値よりも小さい場合、上記処理対象画素の画素値と、上記隣接画素の画素値とに基づいて、当該処理対象画素の補正後の画素値を算出する画素値算出ステップと、を含むことを特徴としている。
本発明に係るフィルタ装置の一実施形態を実施形態1として、図1~14を参照しつつ以下に説明する。本発明に係るフィルタ装置は、複数のブロックに分割して符号化された画像データ(画像情報)を復号する際に生じるブロックノイズを除去するためのものである。このようなブロックノイズは、隣接するブロックの境界において特に顕著である。したがって、本発明に係るフィルタ装置は、ブロックの境界近傍に位置する処理対象画素に対してブロックノイズを除去するためのフィルタ処理を施す。
本実施形態に係るフィルタ装置1の構成を図1に示す。図1は、フィルタ装置1の要部構成を示すブロック図である。図1に示すように、フィルタ装置1は、フィルタ処理前の画素値(補正前の画素値)を入力とし、フィルタ処理後の画素値(補正後の画素値)を出力とする。画素値の入出力の単位は、1画素単位であってもよいし、フィルタ処理の対象となる処理対象画素を有する対象ブロックにおけるブロック境界周辺の複数の画素値を入出力の単位であってもよい。また、フィルタ装置1は、フィルタ処理後の画素値を算出する際に参照する画素値の決定のために、フィルタ処理を施す処理対象画素を有する対象ブロックのサイズを参照する。
ブロックサイズ判定部10は、画像データを分割する複数のブロックにおいて、フィルタ処理を施す処理対象画素を有する対象ブロックのサイズが閾値Th1以上であるか否かを判定する。
差分値算出部11は、画素値の差分絶対値を算出する。より具体的には、対象ブロックに隣接する隣接ブロックの画素に隣接する画素である境界画素の画素値と、隣接ブロックにおいて境界画素に隣接する隣接画素の画素値との差分絶対値を算出する。また、場合によっては、対象ブロックにおける画素の画素値の差分絶対値も算出する。差分値算出部11において算出する差分絶対値は、対象ブロックのサイズに応じて決定される。どのような場合にどのような差分絶対値を算出するのかについては、下記に詳述する。
画素値算出部12は、対象ブロックにおける処理対象画素をフィルタ処理することにより、処理対象画素におけるフィルタ処理後の処理後画素値を算出する。処理対象画素のフィルタ処理前の処理前画素値から処理後画素値を算出する算出方法については、対象ブロックのサイズに応じて決定される。どのような場合にどのような算出方法を用いるのかについては、下記に詳述する。
隣接ブロックサイズ判定部13は、対象ブロックのサイズが閾値Th1より小さいときに、対象ブロックに隣接するいずれかの隣接ブロックのサイズが閾値Th1以上であるか否かを判定する。すなわち、対象ブロックに隣接するブロックに閾値Th1以上のサイズのブロックがあるか否かを判定する。
出力部14は、処理対象画素の処理前画素値を、画素値算出部12において算出した処理後画素値に置き換える。すなわち、処理対象画素の画素値を置換する画素値置換手段である。なお、処理後画素値の出力先については、フィルタ装置1の備えられている装置およびフィルタ装置1の用途に応じて適宜変更される。例えば、フィルタ装置1がループフィルタとして画像復号装置に備えられている場合には、出力部14は復号画像の格納されているフレームメモリに対して出力される。
次に、フィルタ装置1の動作の概要について、図2を参照しつつ以下に説明する。図2は、フィルタ装置1の動作の概要を示すフローチャートである。
フィルタ処理の処理対象画素を有する対象ブロックのサイズが閾値Th1以上である場合における処理対象画素の画素値の算出方法(パターンA)の詳細について、図3~6を参照しつつ以下に説明する。
d=ABS(p0-q0)・・・(1)
を用いて差分絶対値dを算出する。
pk´=(W1×pk+W2×q0)/(W1+W2)・・・(2)
を用いて処理後の画素値pk´を算出する。
p0´=(4×p0+4×q0)/8
p1´=(4×p1+4×q0)/8
p2´=(5×p2+3×q0)/8
p3´=(5×p3+3×q0)/8
p4´=(6×p4+2×q0)/8
p5´=(6×p5+2×q0)/8
p6´=(6×p6+2×q0)/8
p7´=(7×p7+1×q0)/8
となる。
以上説明したように、フィルタ装置1におけるパターンAの処理では、処理対象画素Pkの画素値pkおよび隣接画素Q0の画素値q0の2つの画素の画素値のみから、処理対象画素Pkの処理後の画素値pk´を算出する。これは、ブロックサイズの大きいブロックでは、処理対象画素Pkの周辺の画素の画素値を用いて処理後の画素値pk´を算出したとしても、算出される画素値pk´の値がほとんど変化しないためである。
次に、フィルタ処理の処理対象画素を有する対象ブロックのサイズが閾値Th1未満であり、隣接ブロックのサイズが閾値Th1以上である場合における処理対象画素の画素値の算出方法(パターンB)の詳細について以下に説明する。なお、本項でも、上述したパターンAにおける処理と同様に、フィルタ処理の対象となる画像が図3に示すように分割されており、閾値Th1が「1024」である場合を例に挙げて説明する。すなわち、閾値Th1以上であるブロックは32×32のサイズのブロックであり、閾値Th1未満であるブロックは8×8のサイズおよび4×4のサイズのブロックとなる。
d=ABS(p0-q0) ・・・(3)
ap=ABS(q2-q0)・・・(4)
を用いて差分絶対値dおよび差分絶対値apを算出する。
q0´=(q2+2×q1+2×q0+3×p0)/8
q1´=(q2+q1+q0+p0)/4
q2´=(2×q3+2×q2+q1+q0+2×p0)/8
もちろん、下記式におけるフィルタ係数の数値は一例であり、その数値は適宜変更することができる。
最後に、フィルタ処理の処理対象画素を有する対象ブロックのサイズが閾値Th1未満であり、隣接ブロックのサイズが閾値Th1未満である場合における処理対象画素の画素値の算出方法(パターンC)の詳細について以下に説明する。ここでも、パターンAおよびパターンBの場合と同様に画像データが図3に示すように分割されている場合を例に挙げて説明する。この場合、パターンCにおけるフィルタ処理を施す処理対象画素は、図4および図7において斜線により示されている領域以外の領域に含まれる画素となる。
d=ABS(p0-q0) ・・・(5)
ap=ABS(p2-p0)・・・(6)
を用いて差分絶対値dおよび差分絶対値apを算出する。
p0´=(p2+2×p1+2×p0+2×q0+q1)/8
p1´=(p2+p1+p0+q0)/4
p2´=(2×p3+3×p2+p1+p0+q0)/8
もちろん、下記式におけるフィルタ係数の数値は一例であり、その数値は適宜変更することができる。
q0´=(q2+2×q1+2×q0+2×p0+p1)/8
q1´=(q2+q1+q0+p0)/4
q2´=(2×q3+3×q2+q1+q0+p0)/8
となる。
上述したように、パターンAの処理によるフィルタ方法では、対象ブロックに隣接する隣接ブロックにおいて境界方向に垂直な成分の画素値の変化が大きい場合、すなわち画素値q0周辺の変化が大きい場合にも隣接画素Q0の画素値q0のみを用いる。しかし、対象ブロックに隣接する隣接ブロックにおける隣接画素の画素値において、当該画素値の周辺の変化が大きい場合には、フィルタ処理によって画素値の変化が対象ブロックに及ぼされることがある。この場合、フィルタ処理の対象となる対象ブロックが波打つ(線上のノイズが生じる)ことがある。このような、ノイズは、フィルタ対象範囲が広くなる場合において、特に目立ち、復号した画像が視覚的に不自然に見える可能性が高い。
p0j´=(4×p0j+4×q0j´)/8
p1j´=(4×p1j+4×q0j´)/8
p2j´=(5×p2j+3×q0j´)/8
p3j´=(5×p3j+3×q0j´)/8
p4j´=(6×p4j+2×q0j´)/8
p5j´=(6×p5j+2×q0j´)/8
p6j´=(6×p6j+2×q0j´)/8
p7j´=(7×p7j+1×q0j´)/8
となる。
q0k´´=(5×q0k-1´ + 6×q0k´ + 5×q0k+1´)/16
により算出される画素値q0k´´を画素値q0k´の代わりに用いるようにすることが好ましい。
q0k´´´=(5×q0k-1´´ + 6×q0k´´ + 5×q0k+1´´)/16
を用いればよい。
p0j´=(4×p0j+4×q0j´)/8
p1j´=(4×p1j+4×q0j´)/8
p2j´=(5×p2j+3×q0j´´)/8
p3j´=(5×p3j+3×q0j´´)/8
p4j´=(6×p4j+2×q0j´´´)/8
p5j´=(6×p5j+2×q0j´´´)/8
p6j´=(6×p6j+2×q0j´´´)/8
p7j´=(7×p7j+1×q0j´´´)/8
となる。
次に、閾値Th3、すなわち対象ブロックにおける処理対象画素の範囲の設定について以下に説明する。閾値Th3は、あるブロックにおける行または列に含まれる画素であり、あるブロックに隣接する隣接ブロックに隣接する境界画素から、行または列方向に隣接する画素間の画素値の差分絶対値が最初に閾値Th4(第3の閾値)以上となる画素までの画素数である。閾値Th3は、予め設定されている値であってもよいし、所定のアルゴリズムにより設定される値であってもよい。言い換えれば、フィルタ装置1は、あるブロックの各行または各列に含まれる画素のうち、隣接ブロックに隣接する境界画素から、行方向または列方向に隣接する画素間の画素値の差分絶対値が最初に閾値Th4以上となる画素まで、を処理対象画素とする。
一般的な符号化技術では、符号化対象画像に近い予測画像を生成し、符号化対象画像と予測画像の差である予測誤差を変換し符号化する。したがって、フィルタ処理の対象となる対象ブロックの画素値は、予測画像と予測誤差(逆変換後の信号)の和により復号されたものになる。変換を実行する際のブロックのサイズが大きい場合、予測誤差のブロック内の画素値の変化は小さいことが期待できるが、予測画像に大きな変化があった場合には、予測画像と予測誤差の和により得られる復号画像における画素値の変化が小さいとは限らない。すなわち、予測画像の影響により、復号画像における画素値の変化が大きい場合には、フィルタ処理を施す範囲を小さくする設定する必要がある。
非特許文献1および特許文献1に記載されているフィルタ処理技術においても、境界の状態を示す値として、画素間の画素値の変換に相当する値(例えば、ABS(p2-p0)、ABS(p1-p0))を算出し、この値を用いてフィルタ処理を施すか否かを判定している。しかし、この判定は、あくまで処理対象である境界全体にフィルタ処理を施すか否かを判定するものであり、ブロックの境界からどの程度の画素数離れた画素までフィルタ処理するかを判定するものではない。
次に、閾値Th3の設定方法の変形例について説明する。この変形例は、対象ブロックに隣接する隣接ブロックにおいて境界方向に垂直な成分の画素値の変化が大きい場合、すなわち画素Q0周辺の画素値の変化が大きい場合を考慮した方法である。図8における画素Pijの画素値pijにフィルタ処理する場合の閾値Th3および所定値Tnを、それぞれ閾値Th3j、所定値Tnjとする。
MAGk=ABS(q0k-q0k-1)+ABS(q0k-q0k+1)
を用いて算出する。なお、ここで、kはjの取りえる範囲内における任意の整数である。
MAGk´=(MAGk-1 + 2×MAGk + 2×MAGk+1)/4
を用いて算出されるMAGk´をMAGkの代わりに用いるようにすることが好ましい。
閾値Th3k´=(Th3k-1 + 2×Th3k + 2×Th3k+1)/4
を用いて算出された閾値Th3k´を用いればよい。
上述した説明に用いた所定の閾値Th2、Th4、Tn、α、β、α1、β1、α2、β2、γ1、およびγ2は、量子化パラメータによって変更されることが好ましい。すなわち、量子化パラメータが大きくなるにつれて、閾値も大きくなるように変更されることが好ましい。
非特許文献1および特許文献1に記載されているフィルタ処理技術は、フィルタ処理の対象となる処理対象画素を有する対象ブロックにおける2つのブロック境界(垂直方向の境界および水平方向の境界)を順に処理するものである。しかし、非特許文献1および特許文献1に記載のフィルタ処理技術のように、垂直方向の境界および水平方向の境界を順にフィルタ処理するフィルタ処理技術では、垂直方向の境界および水平方向の境界を同時にフィルタ処理する場合と比較して、復号画像の画質に劣化が生じる。
続いて、平面的なフィルタ処理における処理後の画素値の詳細な算出方法について説明する。なお、本項では、処理対象画素Pijが32×32のサイズのブロックに含まれており、画素Pijの行方向における隣接画素が画素Q0jであり、画素Pijの列方向における隣接画素が画素Ri0である場合を例に挙げて説明する。また、本項では、画素Pijのフィルタ処理前の画素値を画素値pij、画素Q0jの画素値を画素値q0j、画素Ri0の画素値をri0とし、画素Pijのフィルタ処理後の画素値を画素値pij´とする。
pij´=(WP×pij+WV×q0j+WH×ri0)/(WP+WV+WH)・・・(7)
を用いて処理後の画素値pij´を算出する。
WV´=m×WV・・・(8)
WH´=m×WH・・・(9)
WP´=1024-(WV´+WH´)・・・(10)
m=1024/(WP+WV+WH)・・・(11)
を用いて算出することができる。
pij´=(WP´×pij+WV´×q0j+WH´×ri0)>>10・・・(12)
となるように画素値pij´を算出する。
以上説明したように、フィルタ装置1は、平面的なフィルタ処理を施すことにより、処理対象画素に対して水平方向のフィルタ処理と垂直方向のフィルタ処理とが2重に施されてしまうことを防止することができる。
(画像復号装置100の構成)
本発明に係るフィルタ装置を備えた画像復号装置について、実施形態2として図15を参照しつつ以下に説明する。図15は、実施形態2に係る画像復号装置100の要部構成を示すブロック図である。なお、なお、実施形態1と同様の部材に関しては、同一の符号を付し、その説明を省略する。
可変長符号化復号部101は、復号対象ブロックの予測残差データなどの符号化パラメータを表す符号化データを可変長復号処理する。逆量子化・逆変換部102は、入力されたパラメータを逆量子化すると共に、逆変換する。すなわち、逆量子化・逆変換部102は、復号された予測残差データの量子化係数を逆量子化して出力すると共に、予測残差データの変換係数を逆直交変換して出力する。また、ブロックサイズ管理部105は、可変長復号処理により復号された変換サイズを保存する。なお、逆変換における変換方法としては、直交変換が一般的であるが、これに限定されるものではない。例えば、QMFに代表されるフィルタバンクを用いる方法を用いてもよいし、ガボール関数に代表される非直交基底による変換を用いてもよいし、ウェーブレット変換を用いてもよい。
なお、本実施形態では画像復号装置100にフィルタ装置1を備えている場合について例示しているが、もちろん画像符号化装置に備えられているループフィルタとしてフィルタ装置1を備えるようにしてもよい。この場合、フィルタ装置1以外の部材については、従来公知の画像符号化装置と同様の部材を用いることができる。
フィルタ装置1および画像復号装置100に含まれている各ブロックは、ハードウェアロジックによって構成すればよい。または、次のように、CPU(Central Processing Unit)を用いてソフトウェアによって実現してもよい。
10 ブロックサイズ判定部(第1の判定手段)
11 差分値算出部
12 画素値算出部
13 隣接ブロックサイズ判定部(第2の判定手段)
14 出力部
100 画像復号装置
101 可変長符号化復号部
102 逆量子化・逆変換部
103 予測画像生成部
104 フレームメモリ
105 ブロックサイズ管理部
Claims (14)
- サイズの異なる複数のブロックに分割して符号化された画像情報を復号する際に、あるブロックと当該あるブロックに隣接する隣接ブロックとの境界近傍に位置し、かつ、上記あるブロックに属する処理対象画素の画素値を補正するフィルタ装置であって、
上記あるブロックのサイズが第1の閾値以上であるか否かを判定する第1の判定手段と、
上記あるブロックのサイズが上記第1の閾値以上である場合、当該あるブロックに属する少なくとも1つの画素の画素値と、上記隣接ブロックに属する少なくとも1つの画素の画素値とに基づいて、上記処理対象画素の補正後の画素値を算出する画素値算出手段と、
を備えていることを特徴とするフィルタ装置。 - 上記画素値算出手段は、上記あるブロックのサイズが上記第1の閾値以上の場合、当該あるブロックのサイズが上記第1の閾値より小さい場合において用いるよりも少ない数の画素を用いて上記処理対象画素の補正後の画素値を算出することを特徴とする請求項1に記載のフィルタ装置。
- 上記第1の判定手段において、上記あるブロックのサイズが上記第1の閾値以上であると判定された場合、上記処理対象画素を有する行および列の少なくともいずれかに含まれる上記あるブロック内の画素であり、かつ、上記隣接ブロックに隣接する画素である境界画素の画素値と、上記隣接ブロックにおいて当該境界画素に隣接する隣接画素の画素値との差分絶対値を算出する差分値算出手段をさらに備えており、
上記画素値算出手段は、上記差分値算出手段において算出した差分絶対値が第2の閾値よりも小さい場合、上記処理対象画素の画素値と、上記隣接画素の画素値とに基づいて、当該処理対象画素の補正後の画素値を算出することを特徴とする請求項1または2に記載のフィルタ装置。 - 上記画素値算出手段は、上記処理対象画素が、行方向および列方向のいずれの方向においても上記あるブロックと上記隣接ブロックとの境界近傍に位置する場合、上記行方向の境界画素に隣接する隣接画素の画素値、および、上記列方向の境界画素に隣接する隣接画素の画素値を、上記隣接画素の画素値として用いることを特徴とする請求項3に記載のフィルタ装置。
- 上記処理対象画素として設定される画素の範囲は、上記あるブロックのサイズに応じて定められることを特徴とする請求項1から4のいずれか1項に記載のフィルタ装置。
- 上記処理対象画素は、上記あるブロックの各行に含まれる画素の方向に隣接する画素間の画素値の差分絶対値、または、上記あるブロックの各列に含まれる画素の列方向に隣接する画素間の画素値の差分絶対値を用いて設定されることを特徴とする請求項1から4のいずれか1項に記載のフィルタ装置。
- 上記あるブロックに隣接する隣接ブロックのサイズを判定する第2の判定手段をさらに備えており、
上記画素値算出手段は、上記第2の判定手段において判定された隣接ブロックのサイズに応じて、フィルタ強度を変更するように構成されていることを特徴とする請求項1から6のいずれか1項に記載のフィルタ装置。 - 上記あるブロックの各行に含まれる画素のうち、上記隣接ブロックに隣接する境界画素から、行方向に隣接する画素間の画素値の差分絶対値が最初に第3の閾値以上となる画素まで、または、上記あるブロックの各列に含まれる画素のうち、上記隣接ブロックに隣接する境界画素から、列方向に隣接する画素間の画素値の差分絶対値が最初に第3の閾値以上となる画素までを上記処理対象画素とすることを特徴とする請求項6に記載のフィルタ装置。
- 上記第2の判定手段は、上記あるブロックのサイズが上記第1の閾値よりも小さい場合、上記あるブロックに隣接するいずれかの隣接ブロックのサイズが上記第1の閾値以上であるか否かを判定し、
上記画素値算出手段は、上記隣接ブロックのいずれかのサイズが上記第1の閾値以上である場合、上記隣接ブロックのサイズがいずれも上記第1の閾値より小さい場合と比べてフィルタ強度が強くなるように構成されていることを特徴とする請求項7に記載のフィルタ装置。 - 上記画素値算出手段は、下記式(1)を用いて上記処理対象画素の補正後の画素値を算出することを特徴とする請求項3に記載のフィルタ装置。
W1×p+W2×q/(W1+W2)・・・(1)
(上記式(1)中、pは上記処理対象画素の画素値、qは上記隣接画素の画素値、W1およびW2は所定の重み付け係数を示す) - 上記画素値算出手段は、下記式(2)を用いて上記処理対象画素の補正後の画素値を算出することを特徴とする請求項4に記載のフィルタ装置。
WP×p+WH×q+WV×r/(WP+WH+WV)・・・(2)
(上記式(2)中、pは上記処理対象画素の画素値、qは行方向の境界画素に隣接する隣接画素の画素値、rは列方向の境界画素に隣接する隣接画素の画素値、WP、WHおよびWVは所定の重み付け係数を示す) - サイズの異なる複数のブロックに分割して符号化された画像情報を復号する際に、あるブロックと当該あるブロックに隣接する隣接ブロックとの境界近傍に位置し、かつ、上記あるブロックに属する処理対象画素の画素値を補正するフィルタ処理方法であって、
上記あるブロックのサイズが第1の閾値以上であるか否かを判定する第1の判定ステップと、
上記あるブロックのサイズが上記第1の閾値以上である場合、上記処理対象画素を有する行および列の少なくともいずれかに含まれる上記あるブロック内の画素であって、かつ、上記隣接ブロックに隣接する画素である境界画素の画素値と、上記隣接ブロックにおいて当該境界画素に隣接する隣接画素の画素値との差分絶対値を算出する差分値算出ステップと、
算出した上記差分絶対値が第2の閾値よりも小さい場合、上記処理対象画素の画素値と、上記隣接画素の画素値とに基づいて、当該処理対象画素の補正後の画素値を算出する画素値算出ステップと、
を含むことを特徴とするフィルタ処理方法。 - サイズの異なる複数のブロックごとに分割して符号化された画像情報を復号する復号装置であって、請求項1から11のいずれか1項に記載のフィルタ装置を備えていることを特徴とする復号装置。
- 画像情報をサイズの異なる複数のブロックに分割して符号化する符号化装置であって、請求項1から11のいずれか1項に記載のフィルタ装置を備えていることを特徴とする符号化装置。
Priority Applications (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN2009801242043A CN102077590A (zh) | 2008-07-03 | 2009-06-30 | 滤波装置 |
US13/001,054 US20110110603A1 (en) | 2008-07-03 | 2009-06-30 | Filter device |
EP09773489A EP2312853A1 (en) | 2008-07-03 | 2009-06-30 | Filter device |
JP2010519083A JPWO2010001911A1 (ja) | 2008-07-03 | 2009-06-30 | フィルタ装置 |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2008174773 | 2008-07-03 | ||
JP2008-174773 | 2008-07-03 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2010001911A1 true WO2010001911A1 (ja) | 2010-01-07 |
Family
ID=41466003
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2009/062003 WO2010001911A1 (ja) | 2008-07-03 | 2009-06-30 | フィルタ装置 |
Country Status (5)
Country | Link |
---|---|
US (1) | US20110110603A1 (ja) |
EP (1) | EP2312853A1 (ja) |
JP (1) | JPWO2010001911A1 (ja) |
CN (1) | CN102077590A (ja) |
WO (1) | WO2010001911A1 (ja) |
Cited By (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2011125729A1 (ja) * | 2010-04-09 | 2011-10-13 | ソニー株式会社 | 画像処理装置と画像処理方法 |
WO2011145601A1 (ja) * | 2010-05-18 | 2011-11-24 | ソニー株式会社 | 画像処理装置と画像処理方法 |
WO2011145437A1 (ja) * | 2010-05-18 | 2011-11-24 | ソニー株式会社 | 画像処理装置および方法 |
JP2012195702A (ja) * | 2011-03-15 | 2012-10-11 | Fujitsu Ltd | 画像符号化方法及び画像符号化装置 |
CN102948151A (zh) * | 2010-06-17 | 2013-02-27 | 夏普株式会社 | 图像滤波装置、解码装置、编码装置及数据结构 |
JP2013058935A (ja) * | 2011-09-08 | 2013-03-28 | Nippon Hoso Kyokai <Nhk> | 画像処理装置、フィルタ処理方法及びプログラム |
CN103141096A (zh) * | 2010-09-28 | 2013-06-05 | 三星电子株式会社 | 自适应滤波方法和设备 |
CN103765889A (zh) * | 2011-09-09 | 2014-04-30 | 松下电器产业株式会社 | 低复杂度解块滤波器判定 |
JP2017069810A (ja) * | 2015-09-30 | 2017-04-06 | Kddi株式会社 | 動画像の処理装置、処理方法及びコンピュータ可読記憶媒体 |
JP2017076983A (ja) * | 2011-01-18 | 2017-04-20 | ソニー株式会社 | 画像処理装置及び画像処理方法 |
JP2019201438A (ja) * | 2019-09-03 | 2019-11-21 | Kddi株式会社 | 動画像の処理装置、処理方法及びコンピュータ可読記憶媒体 |
WO2019225459A1 (ja) * | 2018-05-23 | 2019-11-28 | パナソニック インテレクチュアル プロパティ コーポレーション オブ アメリカ | 符号化装置、復号装置、符号化方法及び復号方法 |
JP2020036223A (ja) * | 2018-08-30 | 2020-03-05 | Kddi株式会社 | 画像復号装置、画像符号化装置、画像処理システム及びプログラム |
CN114157866A (zh) * | 2016-12-27 | 2022-03-08 | 松下电器(美国)知识产权公司 | 编码装置、解码装置及非暂时性的存储介质 |
Families Citing this family (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102781110B (zh) | 2011-05-12 | 2015-04-01 | 中兴通讯股份有限公司 | 具有跳频功能的资源位置分配方法及装置 |
CN103563374B (zh) * | 2011-05-27 | 2017-02-08 | 索尼公司 | 图像处理设备及方法 |
US9071818B2 (en) * | 2011-08-30 | 2015-06-30 | Organizational Strategies International Pte. Ltd. | Video compression system and method using differencing and clustering |
WO2013059597A1 (en) * | 2011-10-21 | 2013-04-25 | Organizational Strategies International Pte. Ltd. | An interface for use with a video compression system and method using differencing and clustering |
TWI667913B (zh) * | 2012-04-06 | 2019-08-01 | 日商新力股份有限公司 | Image processing device, image processing method, computer program, and recording medium |
RU2737038C2 (ru) * | 2012-06-22 | 2020-11-24 | Сони Корпорейшн | Устройство и способ обработки изображений |
US11310495B2 (en) * | 2016-10-03 | 2022-04-19 | Sharp Kabushiki Kaisha | Systems and methods for applying deblocking filters to reconstructed video data |
US11889054B2 (en) * | 2017-12-29 | 2024-01-30 | Telefonaktiebolaget Lm Ericsson (Publ) | Methods providing encoding and/or decoding of video using reference values and related devices |
CN110290384A (zh) * | 2018-03-19 | 2019-09-27 | 华为技术有限公司 | 图像滤波方法、装置及视频编解码器 |
CN111445427B (zh) * | 2020-05-20 | 2022-03-25 | 青岛信芯微电子科技股份有限公司 | 视频图像处理方法以及显示设备 |
US11770584B1 (en) * | 2021-05-23 | 2023-09-26 | Damaka, Inc. | System and method for optimizing video communications based on device capabilities |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH11205792A (ja) * | 1998-01-14 | 1999-07-30 | Sony Corp | ブロック歪低減方法及び装置 |
JP2001245179A (ja) * | 1999-12-21 | 2001-09-07 | Sanyo Electric Co Ltd | 画像データの歪み低減方法、及びその装置 |
WO2002067589A1 (en) * | 2001-02-23 | 2002-08-29 | Seiko Epson Corporation | Image processing system, image processing method, and image processing program |
JP2003324740A (ja) * | 2002-05-07 | 2003-11-14 | Matsushita Electric Ind Co Ltd | ブロック歪み除去装置 |
JP2004180248A (ja) * | 2001-11-29 | 2004-06-24 | Matsushita Electric Ind Co Ltd | 符号化歪除去方法、動画像符号化方法、動画像復号化方法、およびそれらを実現する装置、プログラム |
JP2005039766A (ja) | 2003-06-27 | 2005-02-10 | Mitsubishi Electric Corp | ポストフィルタ及びポストフィルタ処理方法並びに映像信号復号化装置 |
JP2006521065A (ja) * | 2003-03-17 | 2006-09-14 | クゥアルコム・インコーポレイテッド | 低ビット・レート・ビデオのビデオ品質を改善するための方法及び装置 |
Family Cites Families (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP1469680B1 (en) * | 1996-05-14 | 2008-06-18 | Daewoo Electronics Corporation | Method and apparatus for removing blocking effects in a motion picture decoder |
DE59610133D1 (de) * | 1996-11-28 | 2003-03-20 | Alcatel Sa | Verfahren und Filter zur Reduzierung des Blocking Effektes |
US6317522B1 (en) * | 1998-12-03 | 2001-11-13 | Philips Electronics North America Corp. | Systems and methods for post-processing decompressed images |
US6636645B1 (en) * | 2000-06-29 | 2003-10-21 | Eastman Kodak Company | Image processing method for reducing noise and blocking artifact in a digital image |
US7227901B2 (en) * | 2002-11-21 | 2007-06-05 | Ub Video Inc. | Low-complexity deblocking filter |
JP2005340940A (ja) * | 2004-05-24 | 2005-12-08 | Fuji Photo Film Co Ltd | 画像処理方法及び装置並びにプログラム |
-
2009
- 2009-06-30 EP EP09773489A patent/EP2312853A1/en not_active Withdrawn
- 2009-06-30 CN CN2009801242043A patent/CN102077590A/zh active Pending
- 2009-06-30 JP JP2010519083A patent/JPWO2010001911A1/ja active Pending
- 2009-06-30 WO PCT/JP2009/062003 patent/WO2010001911A1/ja active Application Filing
- 2009-06-30 US US13/001,054 patent/US20110110603A1/en not_active Abandoned
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JPH11205792A (ja) * | 1998-01-14 | 1999-07-30 | Sony Corp | ブロック歪低減方法及び装置 |
JP2001245179A (ja) * | 1999-12-21 | 2001-09-07 | Sanyo Electric Co Ltd | 画像データの歪み低減方法、及びその装置 |
WO2002067589A1 (en) * | 2001-02-23 | 2002-08-29 | Seiko Epson Corporation | Image processing system, image processing method, and image processing program |
JP2004180248A (ja) * | 2001-11-29 | 2004-06-24 | Matsushita Electric Ind Co Ltd | 符号化歪除去方法、動画像符号化方法、動画像復号化方法、およびそれらを実現する装置、プログラム |
JP2003324740A (ja) * | 2002-05-07 | 2003-11-14 | Matsushita Electric Ind Co Ltd | ブロック歪み除去装置 |
JP2006521065A (ja) * | 2003-03-17 | 2006-09-14 | クゥアルコム・インコーポレイテッド | 低ビット・レート・ビデオのビデオ品質を改善するための方法及び装置 |
JP2005039766A (ja) | 2003-06-27 | 2005-02-10 | Mitsubishi Electric Corp | ポストフィルタ及びポストフィルタ処理方法並びに映像信号復号化装置 |
Non-Patent Citations (2)
Title |
---|
"ISO/IEC14496-10", ITU-T, May 2003 (2003-05-01), pages 144 - 153 |
SHINICH SAKAIDA ET AL.: "General Conference", 2006, THE INSTITUTE OF ELECTRONICS, INFORMATION AND COMMUNICATION ENGINEERS, article "Coding of Super-High Definition Video Via Expansion of AVC/H.264 Transform Block Size", pages: D-1 1 - 41 |
Cited By (59)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US9077989B2 (en) | 2010-04-09 | 2015-07-07 | Sony Corporation | Image processing apparatus and image processing method for removing blocking artifacts |
US10373295B2 (en) | 2010-04-09 | 2019-08-06 | Sony Corporation | Image processing apparatus and image processing method |
WO2011125729A1 (ja) * | 2010-04-09 | 2011-10-13 | ソニー株式会社 | 画像処理装置と画像処理方法 |
US9613400B2 (en) | 2010-04-09 | 2017-04-04 | Sony Corporation | Image processing apparatus and image processing method |
CN102823246A (zh) * | 2010-04-09 | 2012-12-12 | 索尼公司 | 图像处理设备和图像处理方法 |
KR102005209B1 (ko) | 2010-05-18 | 2019-07-29 | 소니 주식회사 | 화상 처리 장치와 화상 처리 방법 |
WO2011145437A1 (ja) * | 2010-05-18 | 2011-11-24 | ソニー株式会社 | 画像処理装置および方法 |
US10477206B2 (en) | 2010-05-18 | 2019-11-12 | Sony Corporation | Image processing device and image processing method |
RU2701840C2 (ru) * | 2010-05-18 | 2019-10-01 | Сони Корпорейшн | Устройство и способ обработки изображений |
JPWO2011145601A1 (ja) * | 2010-05-18 | 2013-07-22 | ソニー株式会社 | 画像処理装置と画像処理方法 |
WO2011145601A1 (ja) * | 2010-05-18 | 2011-11-24 | ソニー株式会社 | 画像処理装置と画像処理方法 |
US8855434B2 (en) | 2010-05-18 | 2014-10-07 | Sony Corporation | Image processing device and image processing method |
CN102934429A (zh) * | 2010-05-18 | 2013-02-13 | 索尼公司 | 图像处理装置和图像处理方法 |
US9167267B2 (en) | 2010-05-18 | 2015-10-20 | Sony Corporation | Image processing device and image processing method |
TWI508534B (zh) * | 2010-05-18 | 2015-11-11 | Sony Corp | Image processing apparatus and image processing method |
US9253506B2 (en) | 2010-05-18 | 2016-02-02 | Sony Corporation | Image processing device and image processing method |
KR20180121670A (ko) * | 2010-05-18 | 2018-11-07 | 소니 주식회사 | 화상 처리 장치와 화상 처리 방법 |
KR101914896B1 (ko) | 2010-05-18 | 2018-11-02 | 소니 주식회사 | 화상 처리 장치와 화상 처리 방법 |
KR101808357B1 (ko) | 2010-05-18 | 2017-12-12 | 소니 주식회사 | 화상 처리 장치와 화상 처리 방법 |
CN110430428B (zh) * | 2010-06-17 | 2022-02-11 | 夏普株式会社 | 解码装置、编码装置、解码方法以及编码方法 |
CN102948151A (zh) * | 2010-06-17 | 2013-02-27 | 夏普株式会社 | 图像滤波装置、解码装置、编码装置及数据结构 |
CN106060546A (zh) * | 2010-06-17 | 2016-10-26 | 夏普株式会社 | 解码装置及编码装置 |
CN110430428A (zh) * | 2010-06-17 | 2019-11-08 | 夏普株式会社 | 解码装置、编码装置、解码方法以及编码方法 |
CN103141096A (zh) * | 2010-09-28 | 2013-06-05 | 三星电子株式会社 | 自适应滤波方法和设备 |
US9344716B2 (en) | 2010-09-28 | 2016-05-17 | Samsung Electronics Co., Ltd. | Adaptive filtering method and apparatus |
JP2017076983A (ja) * | 2011-01-18 | 2017-04-20 | ソニー株式会社 | 画像処理装置及び画像処理方法 |
JP2012195702A (ja) * | 2011-03-15 | 2012-10-11 | Fujitsu Ltd | 画像符号化方法及び画像符号化装置 |
JP2013058935A (ja) * | 2011-09-08 | 2013-03-28 | Nippon Hoso Kyokai <Nhk> | 画像処理装置、フィルタ処理方法及びプログラム |
CN107172427A (zh) * | 2011-09-09 | 2017-09-15 | 太阳专利托管公司 | 解码方法及解码装置 |
CN103765889A (zh) * | 2011-09-09 | 2014-04-30 | 松下电器产业株式会社 | 低复杂度解块滤波器判定 |
CN107172427B (zh) * | 2011-09-09 | 2021-01-05 | 太阳专利托管公司 | 解码方法及解码装置 |
CN103765889B (zh) * | 2011-09-09 | 2017-05-24 | 太阳专利托管公司 | 低复杂度解块滤波器判定 |
JP2017069810A (ja) * | 2015-09-30 | 2017-04-06 | Kddi株式会社 | 動画像の処理装置、処理方法及びコンピュータ可読記憶媒体 |
EP3998774A1 (en) * | 2015-09-30 | 2022-05-18 | Kddi Corporation | Moving image processing apparatus, processing method, and computer-readable storage medium |
WO2017056665A1 (ja) * | 2015-09-30 | 2017-04-06 | Kddi株式会社 | 動画像の処理装置、処理方法及びコンピュータ可読記憶媒体 |
EP3358847B1 (en) * | 2015-09-30 | 2021-11-24 | Kddi Corporation | Moving image processing device, processing method and computer-readable storage medium |
CN114173119B (zh) * | 2016-12-27 | 2023-10-20 | 松下电器(美国)知识产权公司 | 解码装置及非暂时性的存储介质 |
CN114157865B (zh) * | 2016-12-27 | 2023-10-20 | 松下电器(美国)知识产权公司 | 编码装置、解码装置及非暂时性的存储介质 |
CN114173118B (zh) * | 2016-12-27 | 2023-10-20 | 松下电器(美国)知识产权公司 | 编码方法、解码方法及发送方法 |
CN114173117B (zh) * | 2016-12-27 | 2023-10-20 | 松下电器(美国)知识产权公司 | 编码方法、解码方法及发送方法 |
CN114157866B (zh) * | 2016-12-27 | 2023-10-20 | 松下电器(美国)知识产权公司 | 编码装置、解码装置及非暂时性的存储介质 |
CN114173119A (zh) * | 2016-12-27 | 2022-03-11 | 松下电器(美国)知识产权公司 | 解码装置及非暂时性的存储介质 |
CN114173117A (zh) * | 2016-12-27 | 2022-03-11 | 松下电器(美国)知识产权公司 | 编码方法、解码方法及发送方法 |
CN114157866A (zh) * | 2016-12-27 | 2022-03-08 | 松下电器(美国)知识产权公司 | 编码装置、解码装置及非暂时性的存储介质 |
CN114157865A (zh) * | 2016-12-27 | 2022-03-08 | 松下电器(美国)知识产权公司 | 编码装置、解码装置及非暂时性的存储介质 |
CN114173118A (zh) * | 2016-12-27 | 2022-03-11 | 松下电器(美国)知识产权公司 | 编码方法、解码方法及发送方法 |
CN112136326A (zh) * | 2018-05-23 | 2020-12-25 | 松下电器(美国)知识产权公司 | 编码装置、解码装置、编码方法和解码方法 |
US11582450B2 (en) | 2018-05-23 | 2023-02-14 | Panasonic Intellectual Property Corporation Of America | Encoder, decoder, encoding method, and decoding method |
WO2019225459A1 (ja) * | 2018-05-23 | 2019-11-28 | パナソニック インテレクチュアル プロパティ コーポレーション オブ アメリカ | 符号化装置、復号装置、符号化方法及び復号方法 |
US11006110B2 (en) | 2018-05-23 | 2021-05-11 | Panasonic Intellectual Property Corporation Of America | Encoder, decoder, encoding method, and decoding method |
JPWO2019225459A1 (ja) * | 2018-05-23 | 2021-04-22 | パナソニック インテレクチュアル プロパティ コーポレーション オブ アメリカPanasonic Intellectual Property Corporation of America | 符号化装置、復号装置、符号化方法及び復号方法 |
CN112136326B (zh) * | 2018-05-23 | 2023-12-12 | 松下电器(美国)知识产权公司 | 编码装置、解码装置、编码方法和解码方法 |
US11856192B2 (en) | 2018-05-23 | 2023-12-26 | Panasonic Intellectual Property Corporation Of America | Encoder, decoder, encoding method, and decoding method |
US11856193B2 (en) | 2018-05-23 | 2023-12-26 | Panasonic Intellectual Property Corporation Of America | Encoder, decoder, encoding method, and decoding method |
US11863743B2 (en) | 2018-05-23 | 2024-01-02 | Panasonic Intellectual Property Corporation Of America | Encoder, decoder, encoding method, and decoding method |
JP6998846B2 (ja) | 2018-08-30 | 2022-01-18 | Kddi株式会社 | 画像復号装置、画像符号化装置、画像処理システム及びプログラム |
WO2020044985A1 (ja) * | 2018-08-30 | 2020-03-05 | Kddi株式会社 | 画像復号装置、画像符号化装置、画像処理システム及びプログラム |
JP2020036223A (ja) * | 2018-08-30 | 2020-03-05 | Kddi株式会社 | 画像復号装置、画像符号化装置、画像処理システム及びプログラム |
JP2019201438A (ja) * | 2019-09-03 | 2019-11-21 | Kddi株式会社 | 動画像の処理装置、処理方法及びコンピュータ可読記憶媒体 |
Also Published As
Publication number | Publication date |
---|---|
EP2312853A1 (en) | 2011-04-20 |
US20110110603A1 (en) | 2011-05-12 |
CN102077590A (zh) | 2011-05-25 |
JPWO2010001911A1 (ja) | 2011-12-22 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2010001911A1 (ja) | フィルタ装置 | |
JP5875979B2 (ja) | フィルタ装置、画像復号装置、画像符号化装置、および、フィルタパラメータのデータ構造 | |
TWI771679B (zh) | 以區塊為基礎之預測技術 | |
RU2696311C1 (ru) | Устройство и способ для компенсации движения видео с выбираемым интерполяционным фильтром | |
US8259819B2 (en) | Method and apparatus for improving video quality by utilizing a unified loop filter | |
JP5684407B2 (ja) | ブロッキングアーチファクトを低減する方法、デブロッキングフィルタユニット、符号器、復号器、ユーザ機器、ネットワーク装置、コンピュータプログラム及びコンピュータプログラム記憶媒体 | |
JP5711098B2 (ja) | 画像符号化方法,画像復号方法,画像符号化装置,画像復号装置およびそれらのプログラム | |
EP2522145B1 (en) | An apparatus and a method for video processing | |
WO2009133844A1 (ja) | エッジを考慮したフィルタリング機能を備えた動画像符号化/復号化方法及び装置 | |
JP5542680B2 (ja) | 画像加工装置、方法及びプログラム、動画像符号化装置、方法及びプログラム、動画像復号装置、方法及びプログラム、並びに、符号化・復号システム及び方法 | |
US20200244965A1 (en) | Interpolation filter for an inter prediction apparatus and method for video coding | |
US8792738B2 (en) | Image decoding apparatus and image coding apparatus | |
JP2024069438A (ja) | イントラ予測を用いた符号化 | |
JP2006295804A (ja) | 画像圧縮方法および画像伸張方法 | |
US11202082B2 (en) | Image processing apparatus and method | |
JP4403094B2 (ja) | スケーラブル符号化方法,スケーラブル復号方法,これらの装置,およびそれらのプログラム,並びにそれらのプログラムを記録したコンピュータ読み取り可能な記録媒体 | |
JP6174966B2 (ja) | 画像符号化装置、画像符号化方法、及びプログラム | |
JPH07307942A (ja) | 画像雑音除去装置 | |
JP4768011B2 (ja) | 動画像符号化装置、及び、動画像復号装置 | |
JP2005260989A (ja) | 画像処理装置及び画像処理方法 | |
JP2009147843A (ja) | 符号化装置及び方法、復号装置及び方法、並びにプログラム | |
JP2009206714A (ja) | デブロッキングフィルタ |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
WWE | Wipo information: entry into national phase |
Ref document number: 200980124204.3 Country of ref document: CN |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 09773489 Country of ref document: EP Kind code of ref document: A1 |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2010519083 Country of ref document: JP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 13001054 Country of ref document: US |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2009773489 Country of ref document: EP |