Nothing Special   »   [go: up one dir, main page]

CN106534871B - The coding method of Video Codec and coding/decoding method - Google Patents

The coding method of Video Codec and coding/decoding method Download PDF

Info

Publication number
CN106534871B
CN106534871B CN201611108487.0A CN201611108487A CN106534871B CN 106534871 B CN106534871 B CN 106534871B CN 201611108487 A CN201611108487 A CN 201611108487A CN 106534871 B CN106534871 B CN 106534871B
Authority
CN
China
Prior art keywords
image
current
reference frame
decoded
encoded
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.)
Active
Application number
CN201611108487.0A
Other languages
Chinese (zh)
Other versions
CN106534871A (en
Inventor
田永鸿
董思维
王耀威
黄铁军
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Peking University
Original Assignee
Peking University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Peking University filed Critical Peking University
Priority to CN201611108487.0A priority Critical patent/CN106534871B/en
Publication of CN106534871A publication Critical patent/CN106534871A/en
Application granted granted Critical
Publication of CN106534871B publication Critical patent/CN106534871B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

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
    • H04N19/50—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
    • H04N19/573—Motion compensation with multiple frame prediction using two or more reference frames in a given prediction direction
    • 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
    • H04N19/13—Adaptive entropy coding, e.g. adaptive variable length coding [AVLC] or context adaptive binary arithmetic coding [CABAC]
    • 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
    • H04N19/57—Motion estimation characterised by a search window with variable size or shape

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Compression Or Coding Systems Of Tv Signals (AREA)

Abstract

The embodiment of the invention provides a kind of coding method of Video Codec and coding/decoding methods.The coding method of the Video Codec, comprising: obtain the corresponding building reference frame of the corresponding former frame encoded image of encoded image of the former frame of current image to be encoded and the current image to be encoded；Update the corresponding building reference frame of the former frame encoded image according to the encoded image, will the corresponding building reference frame of the updated former frame encoded image as the corresponding building reference frame of the corresponding image currently to be encoded of the currently image to be encoded；Using the corresponding building reference frame of the corresponding current image to be encoded of the current image to be encoded as reference picture, the currently image to be encoded is encoded.The present invention can be improved encoding and decoding performance.

Description

The coding method of Video Codec and coding/decoding method
Technical field
The present invention relates to the video coding and decoding technology fields belonged in digital media processing technology more particularly to video to compile solution The coding method of code device and coding/decoding method.
Background technique
Existing method for video coding is usually using the image of encoded mistake as reference frame.In view of existing in cataloged procedure Quantization loss, the not necessarily most reference value of pixel value included in reference frame；Meanwhile the selection of reference frame is not There is the great amount of images information for making full use of front encoded.Therefore, the prediction to current image some regions to be encoded and it is not allowed Really, cause coding efficiency low.
Summary of the invention
The embodiment provides a kind of coding method of Video Codec and coding/decoding methods, can be improved volume solution Code performance.
A kind of coding method of Video Codec, comprising:
Before the encoded image of the former frame of the current image to be encoded of acquisition and the current image to be encoded is corresponding The corresponding building reference frame of one frame encoded image；The building reference frame is the front cross frame according to the current image to be encoded Encoded image and the corresponding forecast image of the front cross frame encoded image generate；
The corresponding building reference frame of the former frame encoded image is updated according to the encoded image, it will be updated The corresponding building reference frame of the former frame encoded image is as the corresponding current figure to be encoded of the current image to be encoded As corresponding building reference frame；
Using the corresponding building reference frame of the corresponding current image to be encoded of the current image to be encoded as reference picture, The current image to be encoded is encoded.
It is described that the corresponding building reference frame of the former frame encoded image is updated according to the encoded image, it will update The corresponding building reference frame of the former frame encoded image afterwards is corresponding currently wait compile as the current image to be encoded The step of code image corresponding building reference frame includes:
According to the relationship between forecast image corresponding to the current encoded image and the current encoded image, Calculate the coded residual of each location of pixels of the current encoded image；
According to the coded residual of each location of pixels of the current encoded image, to the former frame encoded image The pixel value of each location of pixels of corresponding building reference frame is handled, and generates that described current image to be encoded is corresponding works as The corresponding building reference frame of preceding image to be encoded.
The coded residual of each location of pixels according to the current encoded image, it is encoded to the former frame The pixel value of each location of pixels of the corresponding building reference frame of image is handled, and it is corresponding to generate the current image to be encoded Current image to be encoded corresponding building reference frame the step of include:
Step a, judges whether the coded residual of the current pixel position of the current encoded image is less than threshold value；Work as institute State judging result be less than when, then jump to step b, otherwise, jump to step c；
Step b updates the pixel value of the current pixel position of the corresponding building reference frame of the former frame encoded image；
Step c does not update the pixel of the current pixel position of the corresponding building reference frame of the former frame encoded image Value；
According to step a- step c, each location of pixels of building reference frame corresponding to the former frame encoded image Pixel value handled, generate the corresponding building reference frame of current image to be encoded.
Before the step a, the method also includes:
Judge whether the place predicting unit of the current pixel position of the current encoded image uses skip mode to compile Code；
When the judging result, which is, is, then step c is jumped to；Otherwise, step a is jumped to.
The step of current pixel position for updating the corresponding building reference frame of the former frame encoded image includes:
By the pixel value of the current pixel position of the current encoded image and the former frame encoded image Being weighted in corresponding building reference frame with the pixel value of the current pixel position corresponding position, generates and calculates knot Fruit；
Using the calculated result, the current picture of the corresponding building reference frame of the former frame encoded image is updated The pixel value of plain position corresponding position.
When the current image to be encoded is first frame, the image of the former frame of the current image to be encoded is corresponding The value of the corresponding building reference frame of former frame encoded image is set as 0.
The pixel value of the current pixel position by the current encoded image and the former frame are encoded Being weighted in the corresponding building reference frame of image with the pixel value of the current pixel position corresponding position, generates and calculates As a result the step of includes:
P `R(i, j)=PR(i,j)*w1+Prec(i,j)*w2
Wherein, (i, j) is current pixel position, Prec(i, j) is the pixel of the current pixel position of current encoded image Value, PR(i, j) is the pixel value of the current pixel position of the corresponding building reference frame of former frame encoded image；P `R(i, j) is Calculated result；w1With w2Respectively the first weight and the second weight, w1+ w2=1.
A kind of coding/decoding method of Video Codec, comprising:
Before the image of decoding of the former frame of the current image to be decoded of acquisition and the current image to be decoded is corresponding One frame has decoded the corresponding building reference frame of image；The building reference frame is the front cross frame according to the current image to be decoded The corresponding forecast image of the image of decoding for having decoded image and the front cross frame generate；
It has decoded former frame described in image update according to described and has decoded the corresponding building reference frame of image, it will be updated The former frame has decoded the corresponding building reference frame of image as the corresponding building reference frame of the current image to be decoded；
Using the corresponding building reference frame of the current image to be decoded as reference picture, to the currently image to be decoded It is decoded.
It is described to have decoded former frame described in image update according to and decoded the corresponding building reference frame of image, it will update The former frame afterwards has decoded the corresponding building reference frame of image and has referred to as the corresponding building of the current image to be decoded The step of frame includes:
Image and the relationship currently decoded between forecast image corresponding to image have currently been decoded according to described, Calculate the decoded residual of each location of pixels for currently having decoded image；
According to the decoded residual of each location of pixels for currently having decoded image, image has been decoded to the former frame The pixel value of each location of pixels of corresponding building reference frame is handled, and the corresponding structure of the image currently to be decoded is generated Build reference frame.
The decoded residual according to each location of pixels for currently having decoded image, has decoded the former frame The pixel value of each location of pixels of the corresponding building reference frame of image is handled, and it is corresponding to generate the current image to be decoded Building reference frame the step of include:
Step a, judges whether the decoded residual of the current pixel position for currently having decoded image is less than threshold value；Work as institute State judging result be less than when, then jump to step b, otherwise, jump to step c；
Step b updates the pixel value that the former frame has decoded the current pixel position of the corresponding building reference frame of image；
Step c does not update the pixel that the former frame has decoded the current pixel position of the corresponding building reference frame of image Value；
According to step a- step c, each location of pixels of the corresponding building reference frame of image has been decoded to the former frame Pixel value handled, generate the corresponding building reference frame of current image to be decoded.
Before the step a, the method also includes:
Judge whether the place predicting unit of the current pixel position for currently having decoded image uses skip mode solution Code；
When the judging result, which is, is, then step c is jumped to；Otherwise, step a is jumped to.
It is described to update the step of former frame has decoded the current pixel position of the corresponding building reference frame of image and include:
The pixel value of the current pixel position for currently having decoded image and the former frame image has been decoded into Being weighted in corresponding building reference frame with the pixel value of the current pixel position corresponding position, generates and calculates knot Fruit；
Using the calculated result, the current picture that the former frame has decoded the corresponding building reference frame of image is updated The pixel value of plain position corresponding position.
When the current image to be decoded is first frame, the image of the former frame of the current image to be decoded is corresponding The value that former frame has decoded the corresponding building reference frame of image is set as 0.
The pixel value by the current pixel position for currently having decoded image has been decoded with the former frame Being weighted in the corresponding building reference frame of image with the pixel value of the current pixel position corresponding position, generates and calculates As a result the step of includes:
P `R(i, j)=PR(i,j)*w1+Prec(i,j)*w2
Wherein, (i, j) is current pixel position, Prec(i, j) is the pixel for currently having decoded the current pixel position of image Value, PR(i, j) is the pixel value for the current pixel position that former frame has decoded the corresponding building reference frame of image；P `R(i, j) is Calculated result；w1With w2Respectively the first weight and the second weight, w1+ w2=1.
As can be seen from the technical scheme provided by the above-mentioned embodiment of the present invention, in the embodiment of the present invention, reference frame is constructed Selection utilization front encoding and decoding image information, can be improved encoding and decoding performance.
The additional aspect of the present invention and advantage will be set forth in part in the description, these will become from the following description Obviously, or practice through the invention is recognized.
Detailed description of the invention
In order to illustrate the technical solution of the embodiments of the present invention more clearly, required use in being described below to embodiment Attached drawing be briefly described, it should be apparent that, drawings in the following description are only some embodiments of the invention, for this For the those of ordinary skill of field, without any creative labor, it can also be obtained according to these attached drawings other Attached drawing.
Fig. 1 is a kind of schematic diagram of the coding method for Video Codec that the embodiment of the present invention one provides；
Fig. 2 is a kind of schematic diagram of the coding/decoding method for Video Codec that the embodiment of the present invention one provides；
Fig. 3 be the embodiment of the present invention described in Video Codec coding method in construct reference frame generation with it is adaptive Answer the schematic diagram of update method；
Fig. 4 be the embodiment of the present invention described in Video Codec coding/decoding method in construct reference frame generation with it is adaptive Answer the schematic diagram of update method.
Specific embodiment
Embodiments of the present invention are described below in detail, the example of the embodiment is shown in the accompanying drawings, wherein from beginning Same or similar element or element with the same or similar functions are indicated to same or similar label eventually.Below by ginseng The embodiment for examining attached drawing description is exemplary, and for explaining only the invention, and is not construed as limiting the claims.
In order to facilitate understanding of embodiments of the present invention, it is done by taking several specific embodiments as an example below in conjunction with attached drawing further Explanation, and each embodiment does not constitute the restriction to the embodiment of the present invention.
As shown in Figure 1, being a kind of coding method of Video Codec of the present invention, comprising:
Step 11, the encoded image pair of the former frame of current image to be encoded and the current image to be encoded is obtained The corresponding building reference frame of the former frame encoded image answered；The building reference frame is according to the current image to be encoded The corresponding forecast image of the encoded image of the encoded image of front cross frame and the front cross frame generates；When described current to be encoded When image is first frame, the corresponding structure of the corresponding former frame encoded image of image of the former frame of the current image to be encoded The value for building reference frame is set as 0.
Step 12, the corresponding building reference frame of the former frame encoded image is updated according to the encoded image, it will The corresponding building reference frame of the updated former frame encoded image is corresponding current as the current image to be encoded The corresponding building reference frame of image to be encoded；
Step 13, using the corresponding building reference frame of the corresponding current image to be encoded of the current image to be encoded as joining Image is examined, the current image to be encoded is encoded.
In the present invention, the present invention does not use directly a certain encoded image as reference frame, and uses following calculations Method constructs a completely new reference frame, improves encoding characteristics.
Step 12 includes:
Step 121, the forecast image according to corresponding to the current encoded image and the current encoded image it Between relationship, calculate the coded residual of each location of pixels of the current encoded image；Coded residual, that is, encoded image Difference between corresponding forecast image.
Step 122, according to the coded residual of each location of pixels of the current encoded image, to the former frame The pixel value of each location of pixels of the corresponding building reference frame of coded image is handled, and the current image to be encoded is generated The corresponding building reference frame of corresponding current image to be encoded.
In one embodiment, step 122 includes:
Step a, judges whether the coded residual of the current pixel position of the current encoded image is less than threshold value；Work as institute State judging result be less than when, then jump to step b, otherwise, jump to step c；
Step b updates the pixel value of the current pixel position of the corresponding building reference frame of the former frame encoded image；
Step c does not update the pixel of the current pixel position of the corresponding building reference frame of the former frame encoded image Value；
According to step a- step c, each location of pixels of building reference frame corresponding to the former frame encoded image Pixel value handled, generate the corresponding building reference frame of current image to be encoded.
Step b includes:
Step b1 has compiled the pixel value of the current pixel position of the current encoded image with the former frame Being weighted in the corresponding building reference frame of code image with the pixel value of the current pixel position corresponding position, generates meter Calculate result；The pixel value and the encoded figure of the former frame of the current pixel position by the current encoded image As being weighted with the pixel value of the current pixel position corresponding position in corresponding building reference frame, generates and calculate knot The step of fruit includes:
P `R(i, j)=PR(i,j)*w1+Prec(i,j)*w2
Wherein, (i, j) is current pixel position, Prec(i, j) is the pixel of the current pixel position of current encoded image Value, PR(i, j) is the pixel value of the current pixel position of the corresponding building reference frame of former frame encoded image；P `R(i, j) is Calculated result；w1With w2Respectively the first weight and the second weight, w1+ w2=1.
Step b2 updates the institute of the corresponding building reference frame of the former frame encoded image using the calculated result State the pixel value of current pixel position corresponding position.
In another embodiment, step 122 includes:
Step a judges whether the place predicting unit of the current pixel position of the current encoded image uses and skips Pattern-coding；
Step b, judges whether the coded residual of the current pixel position of the current encoded image is less than threshold value；Work as institute State judging result be less than when, then jump to step c, otherwise, jump to step d；
Step c updates the pixel value of the current pixel position of the corresponding building reference frame of the former frame encoded image；
Step d does not update the pixel of the current pixel position of the corresponding building reference frame of the former frame encoded image Value；
According to step a- step d, each location of pixels of building reference frame corresponding to the former frame encoded image Pixel value handled, generate the corresponding building reference frame of current image to be encoded.
As shown in Fig. 2, being a kind of coding/decoding method of Video Codec of the present invention, comprising:
Step 21, the decoding image pair of the former frame of current image to be decoded and the current image to be decoded is obtained The former frame answered has decoded the corresponding building reference frame of image；The building reference frame is according to the current image to be decoded The corresponding forecast image of the image of decoding for having decoded image and the front cross frame of front cross frame generates；When described current to be decoded When image is first frame, the corresponding former frame of image of the former frame of the current image to be decoded has decoded the corresponding structure of image The value for building reference frame is set as 0.
Step 22, it has decoded former frame described in image update according to described and has decoded the corresponding building reference frame of image, it will The updated former frame has decoded the corresponding building reference frame of image as the corresponding building of the current image to be decoded Reference frame；
Step 23, using the corresponding building reference frame of the current image to be decoded as reference picture, to it is described currently to Decoding image is decoded.
Step 22 includes:
Step 221, according to it is described currently decoded image and it is described currently decoded forecast image corresponding to image it Between relationship, calculate it is described currently decoded image each location of pixels decoded residual；
Step 222, according to the decoded residual of each location of pixels for currently having decoded image, to the former frame The pixel value of each location of pixels of the corresponding building reference frame of decoding image is handled, and the current image to be decoded is generated Corresponding building reference frame.
In one embodiment, step 222 includes:
Step a, judges whether the decoded residual of the current pixel position for currently having decoded image is less than threshold value；Work as institute State judging result be less than when, then jump to step b, otherwise, jump to step c；
Step b updates the pixel value that the former frame has decoded the current pixel position of the corresponding building reference frame of image；
Step c does not update the pixel that the former frame has decoded the current pixel position of the corresponding building reference frame of image Value；
According to step a- step c, each location of pixels of the corresponding building reference frame of image has been decoded to the former frame Pixel value handled, generate the corresponding building reference frame of current image to be decoded.
In another embodiment, step 222 includes:
Step a judges whether the place predicting unit of the current pixel position for currently having decoded image uses and skips Mode decoding；When the judging result, which is, is, then step c is jumped to；Otherwise, step a is jumped to.
Step b, judges whether the decoded residual of the current pixel position for currently having decoded image is less than threshold value；Work as institute State judging result be less than when, then jump to step c, otherwise, jump to step d；
Step c updates the pixel value that the former frame has decoded the current pixel position of the corresponding building reference frame of image；
Step d does not update the pixel that the former frame has decoded the current pixel position of the corresponding building reference frame of image Value；
According to step a- step d, each location of pixels of the corresponding building reference frame of image has been decoded to the former frame Pixel value handled, generate the corresponding building reference frame of current image to be decoded.
Step c includes:
Step c1 has solved the pixel value of the current pixel position for currently having decoded image with the former frame Being weighted in the corresponding building reference frame of code image with the pixel value of the current pixel position corresponding position, generates meter Calculate result；The pixel value and former frame decoding figure by the current pixel position for currently having decoded image As being weighted with the pixel value of the current pixel position corresponding position in corresponding building reference frame, generates and calculate knot The step of fruit includes: P `R(i, j)=PR(i,j)*w1+Prec(i,j)*w2；
Wherein, (i, j) is current pixel position, Prec(i,j)For currently decoded image current pixel position pixel Value, PR(i, j) is the pixel value for the current pixel position that former frame has decoded the corresponding building reference frame of image；P `R(i, j) is Calculated result；w1With w2Respectively the first weight and the second weight, w1+ w2=1.
Step c2 updates the institute that the former frame has decoded the corresponding building reference frame of image using the calculated result State the pixel value of current pixel position corresponding position.
Application scenarios of the invention are described below.
The present invention provides the generation and adaptive updates method that reference frame is constructed in a kind of Video Codec.
In an encoding process, firstly, generating building reference frame using encoded image information；Then, building is referred to Each pixel of frame carries out dynamic self-adapting update；Then, the image to be encoded of subsequent input can alternatively be used the building Reference frame carries out predictive coding as reference picture.
Correspondingly, in decoding process, firstly, generating building reference frame using the decoded image information of preamble；Then, Dynamic self-adapting update is carried out to each pixel of building reference frame；Then, the building is chosen to join when subsequent image to be decoded When examining frame, the building reference frame is used to be decoded as reference picture.
The following specifically describes:
The present invention proposes generation and adaptive updates method that reference frame is constructed in a kind of Video Codec.This method packet It includes:
When cataloged procedure starts, building reference frame is generated using encoded image information；
Each pixel value for constructing reference frame is 0；
In an encoding process, according to the relationship between the corresponding forecast image of encoded image, each pixel is analyzed The coded residual of position；
According to coded residual, judge whether the current pixel position for constructing reference frame needs to update；
Building reference frame by update carries out predictive coding as the reference picture of the image to be encoded of subsequent input.
That is, the original pixel values of building reference frame are all 0；With the progress of cataloged procedure, pixel value therein It can constantly update；Update principle are as follows: carry out judging whether to need to update according to coded residual；If updating, new picture is calculated Plain value simultaneously replaces the pixel value that the position is original in building reference frame.
Above-mentioned steps specifically: when cataloged procedure starts, each pixel value for constructing reference frame R is initialized as 0.When the 0th After the completion of frame coding, the reconstruction image of the 0th frame and the relationship of forecast image are analyzed, residual error between the two is obtained.To each pixel The residual error of position is analyzed；When residual absolute value is less than threshold θ, it is believed that the location of pixels for constructing reference frame needs more Newly, it does not otherwise need.For the location of pixels for needing to update in building reference frame, new calculated for pixel values is as follows:
PR(i, j)=PR(i,j)*w1+Prec(i,j)*w2
Wherein, (i, j) is the position of current pixel, PrecFor encoded image (reconstruction image), w1With w2For weight.
It is corresponding, in decoding process, which comprises
When decoding process starts, building each pixel value of reference frame is 0；
When decoding process starts, according to the relationship between the corresponding forecast image of current decoded image, analysis The coded residual of each location of pixels；
According to coded residual, judge to construct whether reference frame current pixel position needs to update；
It is decoded by the building reference frame of update as the reference picture of subsequent image to be decoded.
In a kind of Video Codec proposed by the present invention in the generation of reference frame and adaptive updates method, building is updated The step of reference frame specifically:
In an encoding process, for the location of pixels for needing to update in building reference frame, using corresponding in encoded image The pixel value of position and the pixel value of the position in building reference frame are weighted, and calculated result is as in building reference frame The new pixel value in the position.
Correspondingly, constructing generation and the adaptive updates side of reference frame in a kind of Video Codec proposed by the present invention Method updates the step of constructing reference frame specifically:
In decoding process, for the location of pixels that needs to update in building reference frame, using currently having decoded in image The pixel value of corresponding position and the pixel value of the position in building reference frame are weighted, and calculated result is as building reference The new pixel value in the position in frame.
In the generation and adaptive updates method for constructing reference frame in a kind of Video Codec proposed by the present invention, compiling During code, when predicting unit where the location of pixels of encoded image is using skip mode coding, construct in reference frame The location of pixels is without updating.
Correspondingly, constructing the generation and adaptive updates method of reference frame in a kind of Video Codec proposed by the present invention In, in an encoding process, when predicting unit where currently having decoded the location of pixels of image is using skip mode coding, building The location of pixels in reference frame is without updating.
As shown in figure 3, currently to encode to the 3rd frame, building reference frame R is generated using the 0th, 1,2 frame, and by the 3rd Frame is used for inter-reference.As shown in figure 4, building reference frame is to constantly update, and it makes use of it with the progress of cataloged procedure The information of preceding encoded excessively all images.
The realization platform of the example is HM16.1, and basic configuration is as shown in table 1, and test result is as shown in table 2.
Table 1:
2 this method of table and HM16.1 coding efficiency Comparative result
Sequence sequence Y-component U component V component
Video sequence 1 (576p) - 0.9% - 3.5% - 5.2%
Video sequence 2 (720p) - 1.6% - 2.0% - 2.0%
Video sequence 3 (480p) - 1.3% - 2.3% - 2.9%
Video sequence 4 (1080p) - 1.0% - 0.8% - 0.7%
Average average value - 1.20% - 2.15% - 2.70%
The foregoing is only a preferred embodiment of the present invention, but scope of protection of the present invention is not limited thereto, In the technical scope disclosed by the present invention, any changes or substitutions that can be easily thought of by anyone skilled in the art, It should be covered by the protection scope of the present invention.Therefore, protection scope of the present invention should be with scope of protection of the claims Subject to.

Claims (8)

1.一种视频编解码器的编码方法，其特征在于，包括：1. an encoding method of a video codec, is characterized in that, comprises: 获取当前待编码图像以及所述当前待编码图像的前一帧已编码图像对应的构建参考帧；所述构建参考帧为根据所述当前待编码图像的前两帧已编码图像和所述前两帧已编码图像对应的预测图像产生；Obtain the current image to be encoded and the construction reference frame corresponding to the encoded image of the previous frame of the current image to be encoded; the construction reference frame is based on the first two frames of the current image to be encoded. The predicted image corresponding to the encoded image of the frame is generated; 根据所述已编码图像更新所述前一帧已编码图像对应的构建参考帧，将更新后的所述构建参考帧作为所述当前待编码图像对应的构建参考帧，包括：根据所述当前已编码图像与所述当前已编码图像所对应的预测图像之间的关系，计算所述当前已编码图像的各个像素位置的编码残差；Update the construction reference frame corresponding to the encoded image of the previous frame according to the encoded image, and use the updated construction reference frame as the construction reference frame corresponding to the current image to be encoded, including: the relationship between the coded image and the predicted image corresponding to the currently coded image, and calculating the coding residual of each pixel position of the currently coded image; 根据所述当前已编码图像的各个像素位置的编码残差，对所述前一帧已编码图像对应的构建参考帧的各个像素位置的像素值进行处理，生成所述当前待编码图像对应的构建参考帧；According to the coding residuals of each pixel position of the currently coded image, the pixel values of each pixel position of the construction reference frame corresponding to the previous frame of the coded image are processed to generate the construction corresponding to the current image to be coded. reference frame; 将所述当前待编码图像对应的构建参考帧作为参考图像，对所述当前待编码图像进行编码。The construction reference frame corresponding to the current to-be-encoded image is used as a reference image, and the current to-be-encoded image is encoded. 2.根据权利要求1所述的方法，其特征在于，所述根据所述当前已编码图像的各个像素位置的编码残差，对所述前一帧已编码图像对应的构建参考帧的各个像素位置的像素值进行处理，生成所述当前待编码图像对应的构建参考帧的步骤包括：2 . The method according to claim 1 , wherein, according to the coding residual of each pixel position of the current coded image, each pixel of the reference frame corresponding to the coded image of the previous frame is constructed. 3 . The pixel value of the position is processed, and the step of generating the construction reference frame corresponding to the current to-be-encoded image includes: 步骤a，判断所述当前已编码图像的当前像素位置的编码残差是否小于阈值；当所述判断结果为小于时，则跳到步骤b，否则，跳到步骤c；Step a, judge whether the coding residual of the current pixel position of the current coded image is less than the threshold; when the judgment result is less than, then jump to step b, otherwise, jump to step c; 步骤b，更新所述前一帧已编码图像对应的构建参考帧的当前像素位置的像素值；Step b, update the pixel value of the current pixel position of the construction reference frame corresponding to the encoded image of the previous frame; 步骤c，不更新所述前一帧已编码图像对应的构建参考帧的当前像素位置的像素值；Step c, do not update the pixel value of the current pixel position of the construction reference frame corresponding to the encoded image of the previous frame; 依照步骤a-步骤c，对所述前一帧已编码图像对应的构建参考帧的各个像素位置的像素值进行处理，生成当前待编码图像对应的构建参考帧。According to steps a to c, the pixel values of the respective pixel positions of the construction reference frame corresponding to the encoded image of the previous frame are processed to generate the construction reference frame corresponding to the current to-be-encoded image. 3.根据权利要求2所述的方法，其特征在于，所述步骤a之前，所述方法还包括：3. The method according to claim 2, wherein before the step a, the method further comprises: 判断所述当前已编码图像的当前像素位置的所在预测单元是否采用跳过模式编码；Judging whether the prediction unit where the current pixel position of the currently encoded image is located adopts skip mode encoding; 当所述判断结果为是时，则跳到步骤c；否则，跳到步骤a。When the judgment result is yes, skip to step c; otherwise, skip to step a. 4.根据权利要求2或3所述的方法，其特征在于，所述更新所述前一帧已编码图像对应的构建参考帧的当前像素位置的步骤包括：4. The method according to claim 2 or 3, wherein the step of updating the current pixel position of the construction reference frame corresponding to the encoded image of the previous frame comprises: 将所述当前已编码图像的所述当前像素位置的像素值与所述前一帧已编码图像对应的构建参考帧中的与所述当前像素位置对应处的像素值进行加权计算，生成计算结果；Perform weighted calculation on the pixel value of the current pixel position of the current coded image and the pixel value corresponding to the current pixel position in the construction reference frame corresponding to the coded image of the previous frame, to generate a calculation result ; 使用所述计算结果，更新所述前一帧已编码图像对应的构建参考帧的所述当前像素位置对应处的像素值。Using the calculation result, the pixel value corresponding to the current pixel position of the construction reference frame corresponding to the coded image of the previous frame is updated. 5.一种视频编解码器的解码方法，其特征在于，包括：5. a decoding method of video codec, is characterized in that, comprises: 获取当前待解码图像以及所述当前待解码图像的前一帧的已解码图像对应的前一帧已解码图像对应的构建参考帧；所述构建参考帧为根据所述当前待解码图像的前两帧的已解码图像和所述前两帧的已解码图像对应的预测图像产生；Obtain the current image to be decoded and the construction reference frame corresponding to the decoded image of the previous frame corresponding to the decoded image of the previous frame of the current image to be decoded; the construction reference frame is based on the first two images of the current image to be decoded. generating a predicted image corresponding to the decoded image of the frame and the decoded image of the first two frames; 根据所述已解码图像更新所述前一帧已解码图像对应的构建参考帧，将更新后的所述前一帧已解码图像对应的构建参考帧作为所述当前待解码图像对应的构建参考帧，包括：根据所述当前已解码图像与所述当前已解码图像所对应的预测图像之间的关系，计算所述当前已解码图像的各个像素位置的解码残差；Update the construction reference frame corresponding to the decoded image of the previous frame according to the decoded image, and use the updated construction reference frame corresponding to the decoded image of the previous frame as the construction reference frame corresponding to the current image to be decoded , comprising: calculating, according to the relationship between the currently decoded image and the predicted image corresponding to the currently decoded image, the decoding residuals of each pixel position of the currently decoded image; 根据所述当前已解码图像的各个像素位置的解码残差，对所述前一帧已解码图像对应的构建参考帧的各个像素位置的像素值进行处理，生成所述当前待解码图像对应的构建参考帧；According to the decoding residuals of each pixel position of the current decoded image, the pixel value of each pixel position of the construction reference frame corresponding to the decoded image of the previous frame is processed, and the construction corresponding to the current image to be decoded is generated. reference frame; 将所述当前待解码图像对应的构建参考帧作为参考图像，对所述当前待解码图像进行解码。The construction reference frame corresponding to the current to-be-decoded image is used as a reference image, and the current to-be-decoded image is decoded. 6.根据权利要求5所述的方法，其特征在于，所述根据所述当前已解码图像的各个像素位置的解码残差，对所述前一帧已解码图像对应的构建参考帧的各个像素位置的像素值进行处理，生成所述当前待解码图像对应的构建参考帧的步骤包括：6 . The method according to claim 5 , wherein, according to the decoding residuals of each pixel position of the current decoded image, each pixel of the reference frame corresponding to the decoded image of the previous frame is constructed. 7 . The pixel value of the position is processed, and the step of generating the construction reference frame corresponding to the current image to be decoded includes: 步骤a，判断所述当前已解码图像的当前像素位置的解码残差是否小于阈值；当所述判断结果为小于时，则跳到步骤b，否则，跳到步骤c；Step a, judging whether the decoding residual of the current pixel position of the current decoded image is less than the threshold; when the judgment result is less than, then skip to step b, otherwise, skip to step c; 步骤b，更新所述前一帧已解码图像对应的构建参考帧的当前像素位置的像素值；Step b, update the pixel value of the current pixel position of the construction reference frame corresponding to the decoded image of the previous frame; 步骤c，不更新所述前一帧已解码图像对应的构建参考帧的当前像素位置的像素值；Step c, do not update the pixel value of the current pixel position of the construction reference frame corresponding to the decoded image of the previous frame; 依照步骤a-步骤c，对所述前一帧已解码图像对应的构建参考帧的各个像素位置的像素值进行处理，生成当前待解码图像对应的构建参考帧。According to steps a to c, the pixel values of each pixel position of the construction reference frame corresponding to the decoded image of the previous frame are processed to generate the construction reference frame corresponding to the current image to be decoded. 7.根据权利要求6所述的方法，其特征在于，所述步骤a之前，所述方法还包括：7. The method according to claim 6, wherein before the step a, the method further comprises: 判断所述当前已解码图像的当前像素位置的所在预测单元是否采用跳过模式解码；Determine whether the prediction unit where the current pixel position of the currently decoded image is located adopts skip mode decoding; 当所述判断结果为是时，则跳到步骤c；否则，跳到步骤a。When the judgment result is yes, skip to step c; otherwise, skip to step a. 8.根据权利要求6或7所述的方法，其特征在于，所述更新所述前一帧已解码图像对应的构建参考帧的当前像素位置的步骤包括：8. The method according to claim 6 or 7, wherein the step of updating the current pixel position of the construction reference frame corresponding to the decoded image of the previous frame comprises: 将所述当前已解码图像的所述当前像素位置的像素值与所述前一帧已解码图像对应的构建参考帧中的与所述当前像素位置对应处的像素值进行加权计算，生成计算结果；Perform weighted calculation on the pixel value of the current pixel position of the current decoded image and the pixel value corresponding to the current pixel position in the construction reference frame corresponding to the decoded image of the previous frame to generate a calculation result ; 使用所述计算结果，更新所述前一帧已解码图像对应的构建参考帧的所述当前像素位置对应处的像素值。Using the calculation result, the pixel value corresponding to the current pixel position of the construction reference frame corresponding to the decoded image of the previous frame is updated.
CN201611108487.0A 2016-12-06 2016-12-06 The coding method of Video Codec and coding/decoding method Active CN106534871B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201611108487.0A CN106534871B (en) 2016-12-06 2016-12-06 The coding method of Video Codec and coding/decoding method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201611108487.0A CN106534871B (en) 2016-12-06 2016-12-06 The coding method of Video Codec and coding/decoding method

Publications (2)

Publication Number Publication Date
CN106534871A CN106534871A (en) 2017-03-22
CN106534871B true CN106534871B (en) 2019-06-07

Family

ID=58342693

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201611108487.0A Active CN106534871B (en) 2016-12-06 2016-12-06 The coding method of Video Codec and coding/decoding method

Country Status (1)

Country Link
CN (1) CN106534871B (en)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110235446B (en) * 2017-06-08 2024-04-12 深圳市大疆创新科技有限公司 Video encoding method, video decoding method and related devices
CN109618227B (en) * 2018-10-26 2021-04-20 深圳市野生动物园有限公司 Video data storage method and system
CN112291569B (en) * 2020-06-19 2023-03-24 珠海市杰理科技股份有限公司 Encoding method and device of H.264 image group, encoding and decoding chip and electronic equipment
CN112040234B (en) * 2020-11-04 2021-01-29 北京金山云网络技术有限公司 Video encoding method, video decoding method, video encoding device, video decoding device, electronic equipment and storage medium
CN115633216B (en) * 2022-09-05 2024-05-28 北京智源人工智能研究院 Training method of time domain motion consistency video generation model and video generation method

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6104757A (en) * 1998-05-15 2000-08-15 North Carolina State University System and method of error control for interactive low-bit rate video transmission
CN101232619A (en) * 2008-01-25 2008-07-30 浙江大学 Video Coding Method Embedding Intra-coded Blocks
CN101237587A (en) * 2007-02-02 2008-08-06 中兴通讯股份有限公司 A video sequence coding method and its error control system
CN101355707A (en) * 2008-06-10 2009-01-28 西南交通大学 An error control method for layered coding of stereoscopic video
CN103916667A (en) * 2013-01-07 2014-07-09 华为技术有限公司 Scalable video bit stream encoding and decoding method and device
CN104053009A (en) * 2013-03-15 2014-09-17 中兴通讯股份有限公司 Encoding method of monitoring video and device
CN106488144A (en) * 2016-10-27 2017-03-08 浙江广播电视集团 The system and method that a kind of multi-source information convergence send studio real-time interactive to represent

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8385404B2 (en) * 2008-09-11 2013-02-26 Google Inc. System and method for video encoding using constructed reference frame

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6104757A (en) * 1998-05-15 2000-08-15 North Carolina State University System and method of error control for interactive low-bit rate video transmission
CN101237587A (en) * 2007-02-02 2008-08-06 中兴通讯股份有限公司 A video sequence coding method and its error control system
CN101232619A (en) * 2008-01-25 2008-07-30 浙江大学 Video Coding Method Embedding Intra-coded Blocks
CN101355707A (en) * 2008-06-10 2009-01-28 西南交通大学 An error control method for layered coding of stereoscopic video
CN103916667A (en) * 2013-01-07 2014-07-09 华为技术有限公司 Scalable video bit stream encoding and decoding method and device
CN104053009A (en) * 2013-03-15 2014-09-17 中兴通讯股份有限公司 Encoding method of monitoring video and device
CN106488144A (en) * 2016-10-27 2017-03-08 浙江广播电视集团 The system and method that a kind of multi-source information convergence send studio real-time interactive to represent

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
Representing Visual Objects in HEVC Coding Loop;Tiejun Huang ET AL;《IEEE Circuits and Systems Society 》;20140127;全文
应用于多视点视频压缩的多参考B帧快速编码算法;郁梅;《电子与信息学报》;20080630;全文

Also Published As

Publication number Publication date
CN106534871A (en) 2017-03-22

Similar Documents

Publication Publication Date Title
CN106534871B (en) The coding method of Video Codec and coding/decoding method
US11553175B2 (en) Method and apparatus for candidate list pruning
US9432692B2 (en) Method, apparatus and device for obtaining motion information of video images and template construction method
US20170366817A1 (en) Image coding device and image decoding device
JP2005130509A (en) Intra-prediction method and apparatus
TW201041406A (en) Image predictive encoding device, image predictive encoding method, image predictive encoding program, image predictive decoding device, image predictive decoding method, and image predictive decoding program
US9883200B2 (en) Method of acquiring neighboring disparity vectors for multi-texture and multi-depth video
CN102811346B (en) coding mode selection method and system
CN109076234A (en) Image prediction method and relevant device
CN104853192A (en) Prediction mode selection method and device
CN109688407A (en) Reference block selection method, device, electronic equipment and the storage medium of coding unit
US9374592B2 (en) Mode estimation in pipelined architectures
US8275033B2 (en) Picture mode selection for video transcoding
CN111316646B (en) Candidate MV list construction method and device
CN102984525B (en) A kind of video code flow error concealing method
WO2006100946A1 (en) Image signal re-encoding apparatus and image signal re-encoding method
CN102055987B (en) Error concealment method and device for macroblock subjected to decoding error
CN105323636A (en) Video processing method and video processing device
CN109547788B (en) Image compression method, equipment and image transmission system
CN105992012A (en) Error concealment method and apparatus
CN105812824A (en) Video image encoding method and device
CN109040756A (en) A kind of rapid motion estimating method based on HEVC image content complexity
CN112673630B (en) Bit stream decoder
US11616950B2 (en) Bitstream decoder
CN1993997A (en) Error concealment technique for inter-coded sequences

Legal Events

Date Code Title Description
C06 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