CN112822489A - Hardware implementation method and device for sample adaptive offset compensation filtering - Google Patents
Hardware implementation method and device for sample adaptive offset compensation filtering Download PDFInfo
- Publication number
- CN112822489A CN112822489A CN202011625536.4A CN202011625536A CN112822489A CN 112822489 A CN112822489 A CN 112822489A CN 202011625536 A CN202011625536 A CN 202011625536A CN 112822489 A CN112822489 A CN 112822489A
- Authority
- CN
- China
- Prior art keywords
- boundary
- mode
- compensation
- image block
- value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
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
- H04N19/10—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
- H04N19/146—Data rate or code amount at the encoder output
- H04N19/147—Data rate or code amount at the encoder output according to rate distortion criteria
-
- 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
- H04N19/169—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
-
- 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
- H04N19/186—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 a colour or a chrominance component
-
- 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
- H04N19/50—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
- H04N19/503—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding involving temporal prediction
- H04N19/51—Motion estimation or motion compensation
- H04N19/583—Motion compensation with overlapping blocks
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
The application provides a hardware implementation method and a hardware implementation device for sample adaptive offset compensation filtering, wherein the method comprises the following steps: dividing a source image and a corresponding reconstructed image into a plurality of image blocks according to a preset size; calculating compensation values of the first image block under each color channel in parallel through a compensation filter circuit corresponding to each color channel, wherein the first image block is any image block included in the source image; and compensating the reconstructed image block corresponding to the first image block in the reconstructed image according to the compensation value of the first image block in each color channel. According to the method, the device and the system, sample self-adaptive offset filtering is realized through a hardware circuit, the optimal rate distortion value and the optimal compensation value under each mode are calculated in parallel through the circuit, the optimal compensation value corresponding to the minimum optimal rate distortion value is selected from the optimal rate distortion values under each mode to compensate the reconstructed image, the compensation accuracy is higher, the calculation speed is high, the delay is low, and the real-time requirement of a hardware encoder is met while the 'ringing effect' of the encoder is solved.
Description
Technical Field
The application belongs to the technical field of video processing, and particularly relates to a hardware implementation method and device for sample adaptive offset compensation filtering.
Background
Sample Adaptive Offset (SAO) filtering is pixel compensation of an image to reduce distortion between a source image and a reconstructed image.
At present, a method for realizing sample adaptive offset compensation filtering is provided in the related art, the method performs sample adaptive offset compensation based on a software program, each pixel point in an image needs to be compensated in sequence when the compensation is performed based on the software program, and the pixel-by-pixel compensation mode causes low adaptive offset compensation efficiency, high delay and poor real-time performance.
Disclosure of Invention
The application provides a hardware implementation method and device for sample adaptive offset compensation filtering, the application realizes the sample adaptive offset filtering through a hardware circuit, the optimal rate distortion value and the corresponding optimal compensation value under each mode are calculated through the circuit in parallel, the compensation accuracy is higher, the calculation speed is high, the delay is low, and the real-time requirement of a hardware encoder is met while the ringing effect of the encoder is solved.
An embodiment of a first aspect of the present application provides a hardware implementation method for sample adaptive offset compensation filtering, including:
dividing a source image and a corresponding reconstructed image into a plurality of image blocks according to a preset size;
calculating compensation values of first image blocks under each color channel in parallel through compensation filter circuits corresponding to each color channel, wherein the first image blocks are any image blocks included in the source image;
and compensating a reconstructed image block corresponding to the first image block in the reconstructed image according to the compensation value of the first image block in each color channel.
In some embodiments of the present application, the calculating, in parallel, compensation values of the first image block under each color channel through a compensation filter circuit corresponding to each color channel includes:
inputting a first image block into a first compensation filter circuit corresponding to a first channel, wherein the first channel is any one of color coding YUV (Luma and chroma) channels;
in the first compensation filter circuit, in parallel, a parameter fusion compensation value and a fusion rate distortion value of the first image block under the first channel are calculated through a parameter fusion mode circuit, and a boundary compensation value, a boundary rate distortion value, a stripe compensation value and a stripe rate distortion value of the first image block under the first channel are calculated through a boundary-stripe mode circuit;
and determining a final compensation value of the first image block under the first channel from the parameter fusion compensation value, the boundary compensation value, the strip compensation value and a preset compensation value corresponding to the uncompensated mode according to the fusion rate distortion value, the boundary rate distortion value, the strip rate distortion value and the preset rate distortion value corresponding to the uncompensated mode.
In some embodiments of the present application, the calculating, by the boundary-slice mode circuit, a boundary compensation value, a boundary rate distortion value, a slice compensation value, and a slice rate distortion value for the first image block in the first channel comprises:
calculating, in series by a boundary-to-slice mode circuit, a boundary compensation value and a boundary rate distortion value corresponding to each boundary mode of the first image block in the first channel and a slice compensation value and a slice rate distortion value corresponding to a slice mode, where each boundary mode includes a boundary horizontal mode, a boundary vertical mode, a boundary 45-degree mode, and a boundary 135-degree mode.
In some embodiments of the present application, serially calculating, by a boundary-slice mode circuit, a boundary compensation value and a boundary rate-distortion value corresponding to each boundary mode of the first image block in the first channel includes:
calculating an optimal rate distortion value and a corresponding compensation value corresponding to each type of the first image block under the first channel in parallel according to a first boundary mode by a peak type circuit, a valley type circuit, a reentrant type circuit and a convex type circuit which are included in a boundary-stripe mode circuit, wherein the first boundary mode is any one of the boundary horizontal mode, the boundary vertical mode, the boundary 45-degree mode and the boundary 135-degree mode;
calculating a sum of the optimal rate-distortion values corresponding to the types, and determining the sum as a boundary rate-distortion value of the first image block in the first channel when the first boundary mode is adopted;
and determining the compensation value corresponding to each type as a boundary compensation value of the first image block under the first channel when the first boundary mode is adopted.
In some embodiments of the present application, the calculating, in parallel according to the first boundary mode, an optimal rate-distortion value and a corresponding compensation value for each type of the first image block in the first channel by using a peak type circuit, a valley type circuit, a reentrant type circuit, and a convex type circuit included in the boundary-stripe mode circuit includes:
judging whether each pixel point in a preset number of pixel points in the first image block conforms to a first type in a first boundary mode in parallel through a first type circuit; the first type circuit is any one of a peak type circuit, a valley type circuit, a reentrant type circuit and a salient type circuit which are included in the boundary-stripe mode circuit;
determining all first pixel points which accord with the first type under the first boundary mode from the first image block according to a preset traversal sequence;
and calculating an optimal rate distortion value and a corresponding compensation value of the first image block in the first type under the first boundary mode through the first type circuit according to the number of the first pixel points, the original pixel value and the reconstructed pixel value corresponding to the first pixel points.
In some embodiments of the present application, the determining, according to the fused rate-distortion value, the boundary rate-distortion value, the slice rate-distortion value, and a preset rate-distortion value corresponding to an uncompensated mode, a final compensation value of the first image block in the first channel from the parameter fused compensation value, the boundary compensation value, the slice compensation value, and a preset compensation value corresponding to the uncompensated mode includes:
determining a minimum rate-distortion value from the fusion rate-distortion value, the boundary rate-distortion value, the strip rate-distortion value and a preset rate-distortion value corresponding to a non-compensation mode;
and determining a compensation value corresponding to the minimum rate-distortion value as a final compensation value of the first image block in the first channel.
In some embodiments of the present application, the compensating, according to the compensation value of the first image block in each color channel, a reconstructed image block corresponding to the first image block in the reconstructed image includes:
determining a compensation mode corresponding to a compensation value of the first image block in a first channel;
if the compensation mode is a boundary compensation mode, determining pixel points which respectively accord with a peak type, a valley type, a reentrant type and a salient type in a reconstructed image block corresponding to the first image block; adding the reconstructed pixel values corresponding to the pixel points of each type to the compensation values corresponding to each type under the first channel respectively;
if the compensation mode is a stripe compensation mode, determining pixel points which belong to each stripe in the reconstructed image block respectively; adding the reconstructed pixel value corresponding to the pixel point of each strip to the compensation value corresponding to each strip under the first channel;
and if the compensation mode is a parameter fusion mode, adding the reconstructed pixel values corresponding to the pixel points in the reconstructed image block to the compensation values in the first channel respectively.
An embodiment of the second aspect of the present application provides a hardware implementation apparatus for sample adaptive offset compensation filtering, including:
the image dividing module is used for dividing a source image and a corresponding reconstructed image into a plurality of image blocks according to a preset size;
the three compensation filtering circuits are used for calculating compensation values of the first image block under the YUV three color channels in parallel, and the first image block is any one image block included in the source image;
and the compensation module is used for compensating the reconstructed image block corresponding to the first image block in the reconstructed image according to the compensation values of the first image block under the three color channels.
In some embodiments of the present application, the compensation filtering circuit comprises:
the parameter fusion mode circuit is used for calculating a parameter fusion compensation value and a fusion rate distortion value of the first image block under a first channel, wherein the first channel is any one of color coding YUV channels;
a boundary-slice mode circuit, configured to calculate, by the boundary-slice mode circuit, a boundary compensation value, a boundary rate distortion value, a slice compensation value, and a slice rate distortion value for the first image block in the first channel;
the uncompensated mode circuit is used for providing a preset rate distortion value and a corresponding preset compensation value.
In some embodiments of the present application, the boundary-stripe mode circuitry comprises boundary mode circuitry and stripe mode circuitry, the boundary mode circuitry comprising peak type circuitry, valley type circuitry, and lobe type circuitry;
the slice mode circuit is configured to calculate a slice rate distortion value of the first image block in the first channel and a compensation value corresponding to each slice according to a slice compensation mode;
the boundary mode circuit is configured to calculate, in parallel, an optimal rate distortion value and a corresponding compensation value for each type of the first image block in the first channel when the first boundary mode is adopted through the peak type circuit, the valley type circuit, the reentrant type circuit, and the salient type circuit, and determine a sum of the optimal rate distortion values for each type as a boundary rate distortion value of the first image block in the first channel when the first boundary mode is adopted; and determining the compensation value corresponding to each type as a boundary compensation value of the first image block in the first channel when the first boundary mode is adopted, wherein the first boundary mode is any one of the boundary horizontal mode, the boundary vertical mode, the boundary 45-degree mode and the boundary 135-degree mode.
The technical scheme provided in the embodiment of the application at least has the following technical effects or advantages:
in the embodiment of the application, the sample adaptive offset filtering is realized through a hardware circuit, the optimal rate distortion value and the corresponding optimal compensation value under each mode are calculated in parallel through the circuit, the optimal compensation value corresponding to the minimum optimal rate distortion value is selected from the optimal rate distortion values under each mode to compensate the reconstructed image, the compensation accuracy is higher, the calculation speed is high, the delay is low, and the real-time requirement of a hardware encoder is met while the ringing effect of the encoder is solved.
Additional aspects and advantages of the present application will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the present application.
Drawings
Various other advantages and benefits will become apparent to those of ordinary skill in the art upon reading the following detailed description of the preferred embodiments. The drawings are only for purposes of illustrating the preferred embodiments and are not to be construed as limiting the application. Also, like reference numerals are used to refer to like parts throughout the drawings. In the drawings:
FIG. 1 is a schematic diagram illustrating the partitioning of a boundary pattern provided by an embodiment of the present application;
FIG. 2 is a schematic diagram illustrating shapes of various types of pixel value connecting lines in a boundary mode according to an embodiment of the present application;
FIG. 3 illustrates a schematic view of a strap provided by an embodiment of the present application;
FIG. 4 is a schematic diagram illustrating a left-side neighboring image block and an upper-side neighboring image block referred to by a current image block in a parameter fusion mode according to an embodiment of the present application;
FIG. 5 is a schematic diagram illustrating a 135 degree pattern of the boundary provided by an embodiment of the present application;
FIG. 6 is a flow chart illustrating a hardware implementation of sample adaptive offset compensation filtering according to an embodiment of the present application;
FIG. 7 is a diagram illustrating a hardware circuit structure of sample adaptive offset compensation filtering according to an embodiment of the present application;
FIG. 8 is a schematic diagram illustrating a boundary 45 degree pattern provided by an embodiment of the present application;
FIG. 9 illustrates a schematic diagram of a boundary level pattern provided by an embodiment of the present application;
FIG. 10 shows another schematic of a 135 degree pattern of the boundary as provided by an embodiment of the present application;
FIG. 11 is a schematic diagram illustrating a boundary vertical mode provided by an embodiment of the present application;
fig. 12 is a schematic structural diagram illustrating a hardware implementation apparatus for sample adaptive offset compensation filtering according to an embodiment of the present application.
Detailed Description
Exemplary embodiments of the present application will be described in more detail below with reference to the accompanying drawings. While exemplary embodiments of the present application are shown in the drawings, it should be understood that the present application may be embodied in various forms and should not be limited to the embodiments set forth herein. Rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the scope of the disclosure to those skilled in the art.
It is to be noted that, unless otherwise specified, technical or scientific terms used herein shall have the ordinary meaning as understood by those skilled in the art to which this application belongs.
The following describes a hardware implementation method and apparatus for sample adaptive offset compensation filtering according to an embodiment of the present application with reference to the drawings.
The embodiment of the application provides a hardware implementation method of sample adaptive offset compensation filtering, the compensation modes adopted by the method comprise a boundary mode, a stripe mode and a parameter fusion mode, and rate distortion optimization algorithms are applied in the three modes. Therefore, the boundary mode, the stripe mode, the parameter fusion mode, and the rate-distortion optimization algorithm will be described first.
The boundary mode is a mode of classifying the current pixel by comparing the current pixel value with the adjacent pixel value and compensating the same value for the same type of pixels. Specifically, pixels are divided into 4 patterns according to a position difference EO (Edge Offset) between adjacent pixels. As shown in fig. 1. In fig. 1, a, b, and c are three pixels, respectively, and a pixel c is located between a pixel a and b, and the three pixels are consecutive. The positional relationship between the current pixel and the surrounding pixels is divided into four boundary patterns as shown in fig. 1: EO-0 boundary horizontal mode; EO-90 boundary vertical mode; EO-45 boundary 45 degree mode; EO _135 boundary 135 degree mode.
In any of the above boundary modes, the reconstructed pixels are divided into 5 types, and as shown in table 1, the 5 types are divided according to the magnitude relationship between the pixel value of the pixel point c and the pixel values of the pixel points a and b.
TABLE 1
The pixel position relationship represented by the types 1 to 4 in table 1 is as shown in fig. 2, and the pixel value magnitude relationship shown in table 1 and the pixel value continuous shape corresponding to the types 1 to 4 shown in fig. 2 can be divided into the following 5 types in any one of the boundary horizontal mode, the boundary vertical mode, the boundary 45 degree mode, and the boundary 135 degree mode: valley type, reentrant type, convex type, peak type, and others. Corresponding in turn to types 1, 2, 3, 4 and 0 in table 1 above. In the adaptive offset compensation filtering, the compensation values of the valley type and the valley type are equal to or greater than 0, and the compensation values of the lobe type and the peak type are equal to or less than 0. Other types of compensation values are equal to 0.
The stripe pattern is classified according to the pixel intensity values, the pixel range is equally divided into 32 stripes, then each stripe is compensated according to the characteristic of the pixel of the stripe, and the same compensation value is used for the same stripe. Finally, it is determined by a Rate-Distortion Optimization (Rate-Distortion Optimization) algorithm that 4 different stripes are selected, two of which are continuous and the other two of which are continuous. As shown in fig. 3, the bold 4 bands are the final selected 4 bands.
For each pixel point in the image (represented by 8 bits, with a pixel value range of [0,255]), the way to divide the pixel point into which particular bands is as follows: calculating the pixel value of x which is the pixel point/8; if x is equal to 0, dividing the pixel point into a strip 0; if x is equal to 1, dividing the pixel point into a strip 1; by analogy, …, if x is equal to 31, the pixel point is divided into the bands 31.
The parameter fusion mode is that for an image block, the compensation value of the image block is directly calculated by using the compensation value of the adjacent image block. As shown in fig. 4, the C block may select the already determined compensation values of the a block and the B block, or may use the self-calculated compensation value of the current block. If an image block adjacent to the left or an image block adjacent to the top of the current image block exists, the compensation value already calculated for the image block on the left or the image block on the top can be used to compensate the current image block. In the parameter fusion mode, the compensation value of a specific image block may be calculated by using the above-mentioned boundary mode and/or slice mode.
In the calculation process of the compensation values in the boundary mode, the stripe mode and the parameter fusion mode, a rate-distortion optimization algorithm needs to be used, and specifically, a rate-distortion value is calculated through the following formula (1).
△J＝△D+λR…(1)
In formula (1), Δ D ═ Nh × h-2 hE; λ is a known constant; r is the number of bits coded by the syntax element under the relevant mode type and is relevant to the adopted video coding standard; n represents the number of pixels that fit different types (peak, valley, dip, lobe, 32 stripes) in different boundary patterns (EO _0, EO _90, EO _45, or EO _135), or the number of pixels that belong to different stripes (32 stripes) in a stripe pattern; e represents the sum of the difference values between the original pixel values and the reconstructed pixel values corresponding to all the pixel points which accord with the same type or belong to the same strip; h represents a compensation value, h is E/N, h belongs to the interval [ -7,7], if calculated h is less than-7, h is assigned to-7, if calculated h is greater than 7, h is assigned to 7.
And under the boundary mode, calculating compensation value ranges corresponding to different types in different boundary modes, substituting each compensation value in the range into the range of delta J-delta D + lambda R, and calculating h corresponding to the minimum delta J to be the optimal compensation value.
For example, for mode EO _135, valley class: as shown in fig. 5, for an image block of 64x64 pixels, if the condition c < a & & c < b is satisfied for each pixel compared to the upper left pixel and the lower right pixel, the pixel is divided into the mode EO _135, valley type; the same comparison is performed for all pixels in the 64x64 image block, and it is determined that 20 pixels in total meet the valley type condition, and N is 20. For the determined 20 pixels, the original pixel value of the pixel x (i.e., the pixel value of the pixel x in the source image) and the reconstructed pixel value before the adaptive offset compensation (i.e., the pixel value corresponding to the pixel x in the reconstructed image) are differenced to obtain Ex. Then E-E1 + E2+ … … + Ex + … … + E20. The offset value h must be greater than 0 because of the valley type, and is calculated using h as E/N, and when calculated h is less than 0, the assigned value h is 0, and the offset value range is [0,7 ]. If the calculated h is greater than 7, the assigned value h is 7, and the compensation value range is [0,7 ]. If the calculated h belongs to [0,7], the compensation value range is determined to be [0, h ]. And substituting each integer value in the determined compensation value range into the range of delta J-delta D + lambda R, and obtaining the h value corresponding to the minimum delta J, namely the optimal compensation value h.
Referring to fig. 6, the method specifically includes the following steps:
step 101: and dividing the source image and the corresponding reconstructed image into a plurality of image blocks according to a preset size.
The method comprises the steps of obtaining a source image and a corresponding reconstructed image thereof, and dividing the source image and the corresponding reconstructed image thereof into a plurality of image blocks according to a preset size, wherein the preset size can be 32 × 32 or 64 × 64.
Step 102: and calculating compensation values of the first image block under each color channel in parallel through a compensation filter circuit corresponding to each color channel, wherein the first image block is any image block included in the source image.
The color channels described above represent the luminance Y and chrominance U and V channels in color-coded YUV. The color coding of the source image and the reconstructed image are both YUV format, and if the coding format of the image obtained in step 101 is RGB or RGBA format, it needs to be converted into YUV format first.
The embodiment of the application realizes the sample adaptive offset compensation filtering through a hardware circuit, and the structure of the hardware circuit based on the sample adaptive offset compensation filtering is shown in fig. 7. The circuit structure comprises three compensation filter circuits connected in parallel, and the three compensation filter circuits respectively correspond to Y, U, V channels. The compensation values of the same image block under Y, U, V three channels can be calculated in parallel by the three compensation filter circuits. The parallel computation can improve the computation speed, reduce the delay time of the sample self-adaptive offset compensation filtering, and meet the real-time requirement of a hardware encoder while solving the ringing effect of the encoder.
The hardware structures of the three compensation filter circuits are the same, and fig. 7 only shows the specific structure of the compensation filter circuit in detail. As shown in fig. 7, the compensation filter circuit for the Y channel includes a parallel parameter fusion mode circuit, a boundary-band mode circuit, and an uncompensated mode circuit. The parameter fusion Mode circuit is used for calculating the optimal rate-distortion value and the corresponding compensation value of the image block under the Y channel by adopting a parameter fusion Mode (Merge Mode). The boundary-slice mode circuit is configured to serially employ the boundary mode and the slice mode to calculate an optimal rate-distortion value and a corresponding compensation value for the image block in the Y channel. The uncompensated mode circuit is used for providing a preset rate-distortion value and a corresponding preset compensation value, the preset compensation value is 0, and the preset rate-distortion value is a known initialization constant.
As shown in fig. 7, the boundary-slice mode circuit includes a boundary mode circuit and a slice mode circuit, and the slice mode circuit is configured to calculate an optimal rate-distortion value of the first image block in the Y channel and a compensation value corresponding to each slice according to a slice compensation mode. As shown in fig. 7, the boundary mode circuit includes a peak type circuit, a valley type circuit, a reentrant type circuit, and a convex type circuit, and is configured to calculate, in parallel, an optimal rate-distortion value and a corresponding compensation value corresponding to each type of the image block in the Y channel when an arbitrary boundary mode is adopted by using the peak type circuit, the valley type circuit, the reentrant type circuit, and the convex type circuit. Wherein, the boundary mode may be any one of a boundary horizontal mode, a boundary vertical mode, a boundary 45 degree mode, and a boundary 135 degree mode.
In the embodiment of the present application, the sample adaptive compensation filtering is performed based on the circuit structure shown in fig. 7, and the compensation operation for each image block is the same, so that the embodiment of the present application specifically describes, by taking the first image block as an example, the first image block is any image block included in the source image.
First, a first image block is input into a first compensation filter circuit corresponding to a first channel, wherein the first channel is any one of color coding YUV three channels. Since the circuit structure and the compensation operation of each channel are the same, the embodiment of the present application is specifically described by taking a Y channel as an example, and the specific operations of U and V channels can refer to the Y channel.
Inputting a first image block into a first compensation filter circuit corresponding to a Y channel, calculating a parameter fusion compensation value and a fusion rate distortion value of the first image block under the first channel through a parameter fusion mode circuit in parallel in the first compensation filter circuit, and calculating a boundary compensation value, a boundary rate distortion value, a stripe compensation value and a stripe rate distortion value of the first image block under the first channel through a boundary-stripe mode circuit; and finally, according to the fusion rate distortion value, the boundary rate distortion value, the strip rate distortion value and a preset rate distortion value corresponding to the uncompensated mode, determining a final compensation value of the first image block in the first channel from the parameter fusion compensation value, the boundary compensation value, the strip compensation value and the preset compensation value corresponding to the uncompensated mode.
As shown in fig. 7, the boundary-slice mode circuit includes a boundary mode circuit and a slice mode circuit, the boundary mode circuit serially calculates a boundary compensation value and a boundary rate distortion value corresponding to each boundary mode of the first image block in the first channel, and the slice mode circuit calculates a slice compensation value and a slice rate distortion value corresponding to the first image block in the first channel in the slice mode, each boundary mode includes a boundary horizontal mode, a boundary vertical mode, a boundary 45 degree mode, and a boundary 135 degree mode.
In the embodiment of the present application, the boundary mode circuit and the stripe mode circuit may be two parallel circuits, or may be one circuit, and the calculation of the optimal rate distortion value and the compensation value of the boundary mode and the stripe mode is realized by time division multiplexing the circuits.
The boundary modes comprise a boundary horizontal mode, a boundary vertical mode, a boundary 45-degree mode and a boundary 135-degree mode, and the boundary mode circuit serially calculates the optimal rate-distortion value and the corresponding compensation value of each boundary mode. The processing procedure of each boundary mode is the same, and the first boundary mode is described below as an example, and the first boundary mode is any one of a boundary horizontal mode, a boundary vertical mode, a boundary 45-degree mode, and a boundary 135-degree mode.
As shown in fig. 7, the boundary mode circuit includes a peak type circuit, a valley type circuit, a reentrant type circuit, and a salient type circuit connected in parallel, and the peak type circuit, the valley type circuit, the reentrant type circuit, and the salient type circuit are used to calculate the optimal rate-distortion value and the corresponding compensation value corresponding to each type of the first image block in the first channel in parallel according to the first boundary mode; calculating the sum of the optimal rate-distortion values corresponding to the types, and determining the sum as a boundary rate-distortion value of the first image block in the first channel when the first boundary mode is adopted; and determining the compensation value corresponding to each type as a boundary compensation value of the first image block in the first channel when the first boundary mode is adopted.
The processing procedures of the peak type circuit, the valley type circuit, the reentrant type circuit and the convex type circuit are the same, and the first type circuit is taken as an example and is any one of the peak type circuit, the valley type circuit, the reentrant type circuit and the convex type circuit included in the boundary-stripe mode circuit.
Specifically, whether each pixel point in a preset number of pixel points in the first image block conforms to the first type in the first boundary mode is judged in parallel through the first type circuit. Determining all first pixel points which accord with the first type under the first boundary mode from the first image block according to a preset traversal sequence; and calculating the optimal rate distortion value and the corresponding compensation value of the first image block under the first type in the first boundary mode through the first type circuit according to the number of the first pixel points, the original pixel value and the reconstructed pixel value corresponding to the first pixel points.
For example, in the boundary 45 degree mode shown in fig. 8, the Y-channel value of the current pixel is compared with the Y-channel values of the upper right adjacent pixel and the lower left adjacent pixel, and it is determined which of the 4 types of peak, valley, and lobe the current pixel belongs to. If a or b does not exist in the pixel a/b/c, not operating the pixel c; as shown in fig. 8, since a does not exist in the first row of pixels, no processing is performed on the first row of pixels; the pixels in the left 1 st column, the right 1 st column and the bottom 1 st row are not processed in the boundary 45-degree mode.
When the peak type circuit calculates the number N of the peak type pixels and the difference value E of the corresponding original pixels and the corresponding reconstructed pixels, the class condition judgment is carried out on the pixels of the whole line of the first image block in parallel so as to reduce the calculation time; as shown in fig. 8, it is determined how many pixels in the second row of pixels conform to the peak-like type, and a difference Ex between the Y channel value of the original pixel and the Y channel value of the reconstructed pixel corresponding to the pixels conforming to the peak-like type is calculated; and then adding Ex corresponding to the peak-shaped type pixel to obtain E. The number of pixels of the peak type is accumulated to obtain N. And sequentially carrying out the same calculation on each row from top to bottom, and finally calculating the number N of peak-shaped pixels of the first image block in a boundary 45-degree mode and the difference sum E of the corresponding original pixels and the reconstructed pixels.
In the same operation, the valley type pixel number N. The boundary 45 degree mode will be obtained: a peak type pixel number N0 and a corresponding difference sum E0 of an original pixel and a reconstructed pixel; valley type pixel number N1 and corresponding original pixel and reconstructed pixel difference sum E1; the concave angle type pixel number N2 and the corresponding difference sum of the original pixel and the reconstructed pixel E2; lobe type pixel count N3 and the corresponding original pixel to reconstructed pixel difference sum E3.
And then respectively calculating the optimal rate-distortion value and the corresponding compensation value corresponding to each type through the rate-distortion optimization algorithm. The peak type is exemplified. The compensation value h of the peak type is equal to or less than 0, and is calculated by using h ═ E/N, and when the calculated h is greater than 0, the assigned value h is equal to 0, and the compensation value range is [ -7,0 ]. If the calculated h is less than-7, the assigned value h is-7 and the compensation value range is [ -7,0 ]. If the calculated h belongs to [ -7,0], the compensation value range is determined to be [ h, 0 ]. And substituting each integer value in the determined compensation value range and N0 and E0 corresponding to the peak type into the value delta J, the value delta D and the value lambda R, and obtaining the h value corresponding to the minimum value delta J, namely the optimal compensation value h.
And respectively calculating the optimal rate distortion value and the corresponding compensation value corresponding to the valley type, the reentrant angle type and the convex angle type in the same way as the peak type. And then adding the optimal rate distortion values corresponding to the peak type, the valley type, the reentrant angle type and the convex angle type, determining the obtained sum as the optimal rate distortion value of the first image block in a boundary 45-degree mode, and recording the optimal compensation value h corresponding to the peak type, the valley type, the reentrant angle type and the convex angle type.
For the boundary horizontal mode, the boundary vertical mode and the boundary 135 degree mode, the optimal rate distortion values corresponding to the boundary horizontal mode, the boundary vertical mode and the boundary 135 degree mode and the optimal compensation values corresponding to each type in each mode can be sequentially calculated in the same manner as the boundary 45 degree mode.
It should be noted that, as shown in fig. 9, the leftmost column and the rightmost column of each image block are not processed in the boundary horizontal mode. As shown in fig. 10, the pixels in the 1 st row, the 1 st column on the left side, the 1 st column on the right side and the 1 st row of each image block are not processed in the boundary 135 degree mode. As shown in figure 11 of the drawings,
in the parameter fusion mode circuit, the parameter fusion mode circuit firstly determines whether an image block adjacent to the left side and/or an image block adjacent to the upper side of the first image block exist, if so, and the optimal rate distortion value and the corresponding optimal compensation value of the image block adjacent to the left side and/or the image block adjacent to the upper side are already calculated, the minimum optimal rate distortion value and the corresponding optimal compensation value are selected from the image block adjacent to the left side and/or the image block adjacent to the upper side to be used as the optimal rate distortion value and the optimal compensation value corresponding to the first image block. And if the left adjacent image block and/or the upper adjacent image block do not exist in the source image, calculating the optimal rate-distortion value and the optimal compensation value corresponding to the first image block according to the calculation mode of the boundary mode and/or the strip mode.
After the optimal rate distortion value and the optimal compensation value under each mode are calculated through the method, the minimum rate distortion value is determined from the fusion rate distortion value, the boundary rate distortion value, the strip rate distortion value and the preset rate distortion value corresponding to the uncompensated mode; and determining a compensation value corresponding to the minimum rate-distortion value as a final compensation value of the first image block in the first channel.
For the Y, U, V channels, the optimal rate-distortion value and the optimal compensation value of the first image block under the three channels are calculated by the hardware circuit in the above manner.
Step 103: and compensating the reconstructed image block corresponding to the first image block in the reconstructed image according to the compensation value of the first image block in each color channel.
Determining a compensation mode corresponding to a compensation value of the first image block in the first channel; if the compensation mode is a boundary compensation mode, determining pixel points which respectively accord with a peak type, a valley type, a reentrant angle type and a salient angle type in a reconstructed image block corresponding to the first image block; adding the reconstructed pixel values corresponding to the pixel points of each type to the compensation values corresponding to each type under the first channel respectively; if the compensation mode is a stripe compensation mode, determining pixel points which belong to each stripe in the reconstructed image block respectively; adding the reconstructed pixel value corresponding to the pixel point of each strip to the compensation value corresponding to each strip under the first channel; and if the compensation mode is a parameter fusion mode, adding the reconstructed pixel values corresponding to the pixel points in the reconstructed image block to the compensation values under the first channel respectively.
The hardware circuit is used for operating the pixels with the preset number in parallel, so that the calculation time is reduced, the real-time performance is improved, and the preset number can be one line or multiple lines of pixels. And (4) compensating the filtered pixel, namely reconstructing the pixel and compensating the pixel value, and finally outputting the pixel value after SAO filtering compensation of sample adaptive compensation filtering.
In the embodiment of the application, the sample self-adaptive offset filtering is realized through a hardware circuit, the delay is low, the speed is high, and the real-time requirement of a hardware encoder is met while the ringing effect of the encoder is solved.
The embodiment of the present application further provides a hardware implementation apparatus for sample adaptive offset compensation filtering, where the apparatus is configured to execute the hardware implementation method for sample adaptive offset compensation filtering provided in any of the above embodiments. Referring to fig. 12, the apparatus includes:
the image dividing module 1201 is used for dividing a source image and a corresponding reconstructed image into a plurality of image blocks according to a preset size;
the three compensation filter circuits 1202 are used for calculating compensation values of a first image block under the YUV three color channels in parallel, wherein the first image block is any image block included in the source image;
the compensation module 1203 is configured to compensate a reconstructed image block corresponding to the first image block in the reconstructed image according to the compensation values of the first image block in the three color channels.
As shown in fig. 7, the compensation filter circuit 1202 includes:
the parameter fusion mode circuit is used for calculating a parameter fusion compensation value and a fusion rate distortion value of the first image block under a first channel, wherein the first channel is any one of color coding YUV channels;
the boundary-slice mode circuit is used for calculating a boundary compensation value, a boundary rate distortion value, a slice compensation value and a slice rate distortion value of the first image block under the first channel through the boundary-slice mode circuit;
the uncompensated mode circuit is used for providing a preset rate distortion value and a corresponding preset compensation value.
The boundary-stripe mode circuit comprises a boundary mode circuit and a stripe mode circuit, wherein the boundary mode circuit comprises a peak type circuit, a valley type circuit, a reentrant type circuit and a convex type circuit;
the system comprises a first image block, a slice mode circuit, a slice rate distortion circuit and a slice compensation circuit, wherein the first image block comprises a first channel and a second channel;
the boundary mode circuit is used for calculating the optimal rate distortion value and the corresponding compensation value of each type of the first image block under the first channel when the first boundary mode is adopted in parallel through the peak type circuit, the valley type circuit, the reentrant type circuit and the convex type circuit, and determining the sum of the optimal rate distortion values corresponding to each type as the boundary rate distortion value of the first image block under the first channel when the first boundary mode is adopted; and determining the compensation value corresponding to each type as a boundary compensation value of the first image block in the first channel when a first boundary mode is adopted, wherein the first boundary mode is any one of a boundary horizontal mode, a boundary vertical mode, a boundary 45-degree mode and a boundary 135-degree mode.
The hardware implementation apparatus for sample adaptive offset compensation filtering provided by the foregoing embodiment of the present application and the hardware implementation method for sample adaptive offset compensation filtering provided by the embodiment of the present application are based on the same inventive concept, and have the same beneficial effects as methods adopted, run, or implemented by application programs stored in the hardware implementation apparatus.
It should be noted that:
in the description provided herein, numerous specific details are set forth. However, it is understood that embodiments of the application may be practiced without these specific details. In some instances, well-known structures and techniques have not been shown in detail in order not to obscure an understanding of this description.
Similarly, it should be appreciated that in the foregoing description of exemplary embodiments of the application, various features of the application are sometimes grouped together in a single embodiment, figure, or description thereof for the purpose of streamlining the application and aiding in the understanding of one or more of the various inventive aspects. However, the disclosed method should not be interpreted to reflect the following schematic: this application is intended to cover such departures from the present disclosure as come within known or customary practice in the art to which this invention pertains. Rather, as the following claims reflect, inventive aspects lie in less than all features of a single foregoing disclosed embodiment. Thus, the claims following the detailed description are hereby expressly incorporated into this detailed description, with each claim standing on its own as a separate embodiment of this application.
Furthermore, those skilled in the art will appreciate that while some embodiments described herein include some features included in other embodiments, rather than other features, combinations of features of different embodiments are meant to be within the scope of the application and form different embodiments. For example, in the following claims, any of the claimed embodiments may be used in any combination.
The above description is only for the preferred embodiment of the present application, but the scope of the present application is not limited thereto, and any changes or substitutions that can be easily conceived by those skilled in the art within the technical scope of the present application should be covered within the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
Claims (10)
1. A method for implementing sample adaptive offset compensation filtering in hardware, comprising:
dividing a source image and a corresponding reconstructed image into a plurality of image blocks according to a preset size;
calculating compensation values of first image blocks under each color channel in parallel through compensation filter circuits corresponding to each color channel, wherein the first image blocks are any image blocks included in the source image;
and compensating a reconstructed image block corresponding to the first image block in the reconstructed image according to the compensation value of the first image block in each color channel.
2. The method according to claim 1, wherein the calculating, in parallel, the compensation value of the first image block in each color channel by the compensation filter circuit corresponding to each color channel includes:
inputting a first image block into a first compensation filter circuit corresponding to a first channel, wherein the first channel is any one of color coding YUV (Luma and chroma) channels;
in the first compensation filter circuit, in parallel, a parameter fusion compensation value and a fusion rate distortion value of the first image block under the first channel are calculated through a parameter fusion mode circuit, and a boundary compensation value, a boundary rate distortion value, a stripe compensation value and a stripe rate distortion value of the first image block under the first channel are calculated through a boundary-stripe mode circuit;
and determining a final compensation value of the first image block under the first channel from the parameter fusion compensation value, the boundary compensation value, the strip compensation value and a preset compensation value corresponding to the uncompensated mode according to the fusion rate distortion value, the boundary rate distortion value, the strip rate distortion value and the preset rate distortion value corresponding to the uncompensated mode.
3. The method of claim 2, wherein the calculating, by the boundary-slice mode circuit, the boundary compensation value, the boundary rate distortion value, the slice compensation value, and the slice rate distortion value for the first image block in the first channel comprises:
calculating, in series by a boundary-to-slice mode circuit, a boundary compensation value and a boundary rate distortion value corresponding to each boundary mode of the first image block in the first channel and a slice compensation value and a slice rate distortion value corresponding to a slice mode, where each boundary mode includes a boundary horizontal mode, a boundary vertical mode, a boundary 45-degree mode, and a boundary 135-degree mode.
4. The method as claimed in claim 3, wherein serially calculating boundary compensation values and boundary rate distortion values corresponding to boundary modes of the first image block in the first channel by a boundary-slice mode circuit comprises:
calculating an optimal rate distortion value and a corresponding compensation value corresponding to each type of the first image block under the first channel in parallel according to a first boundary mode by a peak type circuit, a valley type circuit, a reentrant type circuit and a convex type circuit which are included in a boundary-stripe mode circuit, wherein the first boundary mode is any one of the boundary horizontal mode, the boundary vertical mode, the boundary 45-degree mode and the boundary 135-degree mode;
calculating a sum of the optimal rate-distortion values corresponding to the types, and determining the sum as a boundary rate-distortion value of the first image block in the first channel when the first boundary mode is adopted;
and determining the compensation value corresponding to each type as a boundary compensation value of the first image block under the first channel when the first boundary mode is adopted.
5. The method according to claim 4, wherein the calculating the optimal rate-distortion value and the corresponding compensation value for each type of the first image block under the first channel in parallel according to the first boundary mode by the peak type circuit, the valley type circuit, the reentrant type circuit and the convex type circuit included in the boundary-stripe mode circuit comprises:
judging whether each pixel point in a preset number of pixel points in the first image block conforms to a first type in a first boundary mode in parallel through a first type circuit; the first type circuit is any one of a peak type circuit, a valley type circuit, a reentrant type circuit and a salient type circuit which are included in the boundary-stripe mode circuit;
determining all first pixel points which accord with the first type under the first boundary mode from the first image block according to a preset traversal sequence;
and calculating an optimal rate distortion value and a corresponding compensation value of the first image block in the first type under the first boundary mode through the first type circuit according to the number of the first pixel points, the original pixel value and the reconstructed pixel value corresponding to the first pixel points.
6. The method according to claim 2, wherein the determining, according to the fused rate-distortion value, the boundary rate-distortion value, the slice rate-distortion value, and a preset rate-distortion value corresponding to an uncompensated mode, a final compensation value of the first image block in the first channel from the parameter fused compensation value, the boundary compensation value, the slice compensation value, and a preset compensation value corresponding to the uncompensated mode comprises:
determining a minimum rate-distortion value from the fusion rate-distortion value, the boundary rate-distortion value, the strip rate-distortion value and a preset rate-distortion value corresponding to a non-compensation mode;
and determining a compensation value corresponding to the minimum rate-distortion value as a final compensation value of the first image block in the first channel.
7. The method according to claim 1, wherein the compensating the reconstructed image block corresponding to the first image block in the reconstructed image according to the compensation value of the first image block in each color channel comprises:
determining a compensation mode corresponding to a compensation value of the first image block in a first channel;
if the compensation mode is a boundary compensation mode, determining pixel points which respectively accord with a peak type, a valley type, a reentrant type and a salient type in a reconstructed image block corresponding to the first image block; adding the reconstructed pixel values corresponding to the pixel points of each type to the compensation values corresponding to each type under the first channel respectively;
if the compensation mode is a stripe compensation mode, determining pixel points which belong to each stripe in the reconstructed image block respectively; adding the reconstructed pixel value corresponding to the pixel point of each strip to the compensation value corresponding to each strip under the first channel;
and if the compensation mode is a parameter fusion mode, adding the reconstructed pixel values corresponding to the pixel points in the reconstructed image block to the compensation values in the first channel respectively.
8. An apparatus for sample adaptive offset compensation filtering, comprising:
the image dividing module is used for dividing a source image and a corresponding reconstructed image into a plurality of image blocks according to a preset size;
the three compensation filtering circuits are used for calculating compensation values of the first image block under the YUV three color channels in parallel, and the first image block is any one image block included in the source image;
and the compensation module is used for compensating the reconstructed image block corresponding to the first image block in the reconstructed image according to the compensation values of the first image block under the three color channels.
9. The apparatus of claim 8, wherein the compensation filtering circuit comprises:
the parameter fusion mode circuit is used for calculating a parameter fusion compensation value and a fusion rate distortion value of the first image block under a first channel, wherein the first channel is any one of color coding YUV channels;
a boundary-slice mode circuit, configured to calculate, by the boundary-slice mode circuit, a boundary compensation value, a boundary rate distortion value, a slice compensation value, and a slice rate distortion value for the first image block in the first channel;
the uncompensated mode circuit is used for providing a preset rate distortion value and a corresponding preset compensation value.
10. The apparatus of claim 9, wherein the boundary-stripe mode circuit comprises a boundary mode circuit and a stripe mode circuit, the boundary mode circuit comprising a peak type circuit, a valley type circuit, and a lobe type circuit;
the slice mode circuit is configured to calculate a slice rate distortion value of the first image block in the first channel and a compensation value corresponding to each slice according to a slice compensation mode;
the boundary mode circuit is configured to calculate, in parallel, an optimal rate distortion value and a corresponding compensation value for each type of the first image block in the first channel when the first boundary mode is adopted through the peak type circuit, the valley type circuit, the reentrant type circuit, and the salient type circuit, and determine a sum of the optimal rate distortion values for each type as a boundary rate distortion value of the first image block in the first channel when the first boundary mode is adopted; and determining the compensation value corresponding to each type as a boundary compensation value of the first image block in the first channel when the first boundary mode is adopted, wherein the first boundary mode is any one of the boundary horizontal mode, the boundary vertical mode, the boundary 45-degree mode and the boundary 135-degree mode.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011625536.4A CN112822489B (en) | 2020-12-30 | 2020-12-30 | Hardware implementation method and device for sample self-adaptive offset compensation filtering |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011625536.4A CN112822489B (en) | 2020-12-30 | 2020-12-30 | Hardware implementation method and device for sample self-adaptive offset compensation filtering |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112822489A true CN112822489A (en) | 2021-05-18 |
| CN112822489B CN112822489B (en) | 2023-05-16 |
Family
ID=75854807
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011625536.4A Active CN112822489B (en) | 2020-12-30 | 2020-12-30 | Hardware implementation method and device for sample self-adaptive offset compensation filtering |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112822489B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113645466A (en) * | 2021-06-29 | 2021-11-12 | 深圳市迪威码半导体有限公司 | Image removal block based on random probability |
| WO2022022176A1 (en) * | 2020-07-30 | 2022-02-03 | 华为技术有限公司 | Image processing method and related device |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130051455A1 (en) * | 2011-08-24 | 2013-02-28 | Vivienne Sze | Flexible Region Based Sample Adaptive Offset (SAO) and Adaptive Loop Filter (ALF) |
| WO2013109419A1 (en) * | 2012-01-21 | 2013-07-25 | General Instrument Corporation | Devices and methods for sample adaptive offset coding and/or selection of band offset parameters |
| US20140341287A1 (en) * | 2013-05-20 | 2014-11-20 | Texas Instruments Incorporated | Sample Adaptive Offset (SAO) Filtering in Video Coding |
| CN104823446A (en) * | 2012-12-06 | 2015-08-05 | 索尼公司 | Image processing device, image processing method, and program |
| CN105409221A (en) * | 2014-04-29 | 2016-03-16 | 微软技术许可有限责任公司 | Encoder-side decisions for sample adaptive offset filtering |
| US20170006283A1 (en) * | 2015-06-30 | 2017-01-05 | Microsoft Technology Licensing, Llc | Computationally efficient sample adaptive offset filtering during video encoding |
-
2020
- 2020-12-30 CN CN202011625536.4A patent/CN112822489B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130051455A1 (en) * | 2011-08-24 | 2013-02-28 | Vivienne Sze | Flexible Region Based Sample Adaptive Offset (SAO) and Adaptive Loop Filter (ALF) |
| WO2013109419A1 (en) * | 2012-01-21 | 2013-07-25 | General Instrument Corporation | Devices and methods for sample adaptive offset coding and/or selection of band offset parameters |
| CN104823446A (en) * | 2012-12-06 | 2015-08-05 | 索尼公司 | Image processing device, image processing method, and program |
| US20140341287A1 (en) * | 2013-05-20 | 2014-11-20 | Texas Instruments Incorporated | Sample Adaptive Offset (SAO) Filtering in Video Coding |
| CN105409221A (en) * | 2014-04-29 | 2016-03-16 | 微软技术许可有限责任公司 | Encoder-side decisions for sample adaptive offset filtering |
| US20170006283A1 (en) * | 2015-06-30 | 2017-01-05 | Microsoft Technology Licensing, Llc | Computationally efficient sample adaptive offset filtering during video encoding |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2022022176A1 (en) * | 2020-07-30 | 2022-02-03 | 华为技术有限公司 | Image processing method and related device |
| CN113645466A (en) * | 2021-06-29 | 2021-11-12 | 深圳市迪威码半导体有限公司 | Image removal block based on random probability |
| CN113645466B (en) * | 2021-06-29 | 2024-03-08 | 深圳市迪威码半导体有限公司 | Image deblocking algorithm based on random probability |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112822489B (en) | 2023-05-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| KR100791375B1 (en) | Apparatus and method for color correction | |
| US5325186A (en) | Motion adaptive luminance signal and color signal separating filter | |
| EP3082339B1 (en) | Apparatus and method of offset correction for video coding | |
| EP1746846B1 (en) | Image processing device having color shift-correcting function, image processing program, and electronic camera | |
| US9256926B2 (en) | Use of inpainting techniques for image correction | |
| US7095889B2 (en) | Method of and apparatus for image processing | |
| US7956871B2 (en) | Color disparity correction in image sensors methods and circuits | |
| KR100574185B1 (en) | Color sample interpolation | |
| CN112822489A (en) | Hardware implementation method and device for sample adaptive offset compensation filtering | |
| US7812868B2 (en) | Signal processing method and apparatus, and image sensing apparatus | |
| KR100332268B1 (en) | How to make hierarchical estimates of movement in television signals | |
| US5999230A (en) | Tone correction apparatus for digital color image signal | |
| JP3862374B2 (en) | Image processing apparatus and method | |
| KR100398564B1 (en) | Solid-state color imager | |
| US6967748B1 (en) | Color image pickup device | |
| US20090010561A1 (en) | Device for removing noise in image data | |
| US8253817B2 (en) | Image compression method, device, electronic camera, and program | |
| CN114007074B (en) | Method and device for screening intra-frame prediction modes of video frames | |
| US6826302B2 (en) | Luminance correction of colorless low saturation regions using correction constants calculated from color saturation values | |
| US20020101434A1 (en) | Image interpolating device | |
| CN108259903B (en) | H.265 sample self-adaptive compensation method based on human eye region of interest | |
| US11379956B2 (en) | Image processing circuit and associated image processing method | |
| US11748852B2 (en) | Pixel interpolation device and pixel interpolation method, and image processing device, and program and recording medium | |
| JP2004159147A (en) | Block noise reduction circuit | |
| CN112351152B (en) | Image processing circuit and related image processing method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |