WO2009086761A1 - 视频编码、解码方法及装置和视频处理*** - Google Patents

视频编码、解码方法及装置和视频处理*** Download PDF

Info

Publication number
WO2009086761A1
WO2009086761A1 PCT/CN2008/073291 CN2008073291W WO2009086761A1 WO 2009086761 A1 WO2009086761 A1 WO 2009086761A1 CN 2008073291 W CN2008073291 W CN 2008073291W WO 2009086761 A1 WO2009086761 A1 WO 2009086761A1
Authority
WO
WIPO (PCT)
Prior art keywords
information
macroblock
reference image
offset information
offset
Prior art date
Application number
PCT/CN2008/073291
Other languages
English (en)
French (fr)
Inventor
Haitao Yang
Sixin Lin
Shan Gao
Yingjia Liu
Jiali Fu
Jiantong Zhou
Original Assignee
Huawei Technologies Co., Ltd.
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 Huawei Technologies Co., Ltd. filed Critical Huawei Technologies Co., Ltd.
Publication of WO2009086761A1 publication Critical patent/WO2009086761A1/zh
Priority to US12/830,126 priority Critical patent/US20100266048A1/en

Links

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/50Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
    • H04N19/503Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding involving temporal prediction
    • H04N19/51Motion estimation or motion compensation
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/50Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
    • H04N19/597Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding specially adapted for multi-view video sequence encoding
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/60Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using transform coding
    • H04N19/61Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using transform coding in combination with predictive coding

Definitions

  • Embodiments of the present invention relate to the field of video technologies, and in particular, to a video encoding method and apparatus, a video decoding method and apparatus, and a video processing system. Background technique
  • MSM Inter-predicted motion information skip mode
  • MVC Multi-view Video Coding
  • the MSM technology mainly includes two steps of calculating global disparity vector information (Global Disparity Vector; hereinafter referred to as GDV) and calculating motion information of corresponding macroblocks in the reference image.
  • GDV Global Disparity Vector
  • the upper and lower squares on both sides represent an anchor picture in an adjacent view, and there may be multiple non-anchor frames between the anchor frame Img A and the anchor frame Img B. Only one non-anchor frame Img eur is shown in Fig.
  • each macroblock in the non-anchor frame Img CM is a corresponding macroblock in the inter-view reference picture, for example, the macroblock MB CM in the figure is a corresponding macro block in the inter-view reference picture.
  • the overhead of RDCost MBcM , MSM if the MSM mode overhead is less than the other macroblock mode overhead, the MSM is selected as the final mode of the macroblock.
  • determining, by using the GDV cur information, that the corresponding macroblock may not be the corresponding macroblock that optimizes the current macroblock coding efficiency may be preset in the reference image in order to find a corresponding macroblock that optimizes the current macroblock coding efficiency.
  • the motion information of the corresponding macroblock that optimizes the current macroblock coding efficiency is sought, and the motion information of the current macroblock is obtained.
  • each block in the search range is searched by using an index identifier, and the index numbers are 0, 1, 2, 3, and so on.
  • the current coding is performed.
  • the macro block MB is simultaneously encoded with the index number "5".
  • the index number encoding method in the method is one-dimensional positional deviation.
  • the shifting information does not reveal the statistical characteristics of the positional offset information in the horizontal direction and the vertical direction, which affects the coding efficiency.
  • the motion information of the corresponding macroblock pointed by the GDV information in the forward-view or backward-view reference image is used as the motion information of the current coded macroblock, and is used for motion compensation of the currently coded macroblock, but
  • the difference in the corresponding macroblock in the reference picture of the view or the backward view makes the coding efficiency low.
  • the embodiment of the invention provides a video encoding method and device, and a video decoding method and device
  • the video processing system is used to solve the defect of low coding efficiency in the prior art, and achieve high-efficiency coding of video images.
  • the embodiment of the invention provides a video encoding method, including:
  • the first offset information is encoded.
  • the embodiment of the invention provides a video decoding method, including:
  • An embodiment of the present invention provides a video encoding apparatus, including:
  • a first module configured to obtain, according to a predetermined search accuracy, disparity vector information, an image block corresponding to the current macroblock and having the same size as the search precision in the reference image of the adjacent view; Obtaining, according to a coordinate system established by the image block, first offset information of a corresponding macroblock that makes the current macroblock encoding efficiency optimal;
  • a third module configured to encode the first offset information.
  • An embodiment of the present invention provides a video decoding apparatus, including:
  • a fifth module configured to parse the received code stream information, and obtain a parameter adjacent to the current macro block a first offset information of a macroblock corresponding to the current macroblock
  • a sixth module configured to obtain an image block corresponding to the current macroblock in the reference image of the adjacent view according to the disparity vector information ;
  • a seventh module configured to obtain coordinate information of a macroblock corresponding to the current macroblock according to the first offset information in a coordinate system of a reference image search area established according to the image block;
  • an eighth module configured to obtain motion information of the macroblock corresponding to the current macroblock according to the coordinate information, and apply the motion information to perform motion compensation.
  • An embodiment of the present invention provides a video processing system, including a video encoding apparatus and a video decoding apparatus, where the video encoding apparatus includes:
  • a first module configured to obtain, according to a predetermined search accuracy, disparity vector information, an image block corresponding to the current macroblock and having the same size as the search precision in the reference image of the adjacent view; Obtaining, according to a coordinate system established by the image block, first offset information of a corresponding macroblock that makes the current macroblock encoding efficiency optimal;
  • a third module configured to encode the first offset information
  • the video decoding device includes:
  • a fifth module configured to parse the received code stream information, to obtain first offset information of a macroblock corresponding to the current macroblock in a reference image adjacent to the current macroblock;
  • a sixth module configured to obtain, according to the disparity vector information, an image block corresponding to the current macroblock in the reference image of the adjacent view;
  • a seventh module configured to obtain coordinate information of a macroblock corresponding to the current macroblock according to the first offset information in a coordinate system of a reference image search area established according to the image block;
  • an eighth module configured to obtain motion information of the macroblock corresponding to the current macroblock according to the coordinate information, and apply the motion information to perform motion compensation.
  • the embodiment of the invention provides a video encoding method, including: XORing the mark symbol of the current macroblock for marking the forward and backward view with the mark symbol of one or more macro blocks around it;
  • a context model is established based on the tag symbols of the one or more macroblocks of the periphery, and the XOR-processed tag symbol information is encoded using the context model.
  • the video encoding method and apparatus provided by the embodiments of the present invention, the video decoding method and apparatus, and the video processing system establish the vertical and horizontal coordinate position information of each block in the search area by selecting the coordinate origin of the appropriate search area;
  • the information of the block peripheral block as a context for encoding the position offset information of the corresponding macroblock in the reference picture of the adjacent macroblock, encodes the offset of the current macroblock, and improves the coding efficiency.
  • FIG. 1 is a schematic diagram of a conventional GDV derivation coding process
  • FIG. 2 is a schematic diagram of encoding processing of location information within a range of existing search areas
  • Embodiment 3 is a flowchart of Embodiment 1 of a video encoding method according to the present invention.
  • FIG. 4 is a schematic diagram of coordinate origin selection and offset amount encoding processing of a search area in Embodiment 2 of the video encoding method of the present invention
  • FIG. 5 is a schematic diagram of offset coordinate coding of a corresponding macroblock of a current macroblock in Embodiment 2 of a video encoding method according to the present invention
  • FIG. 6 is a schematic diagram of coordinate origin selection and offset amount encoding processing of a search area in Embodiment 3 of the video encoding method of the present invention.
  • FIG. 7 is a flowchart of an embodiment of a video decoding method according to the present invention.
  • Embodiment 8 is a schematic structural diagram of Embodiment 1 of a video encoding apparatus according to the present invention.
  • Embodiment 9 is a schematic structural diagram of Embodiment 2 of a video encoding apparatus according to the present invention.
  • FIG. 10 is a schematic structural diagram of Embodiment 1 of a video decoding apparatus according to the present invention.
  • Embodiment 11 is a schematic structural diagram of Embodiment 2 of a video decoding apparatus according to the present invention.
  • FIG. 12 is a schematic structural diagram of Embodiment 1 of a video processing system according to the present invention
  • FIG. 13 is a schematic structural diagram of Embodiment 2 of a video processing system according to the present invention.
  • FIG. 3 is a flowchart of Embodiment 1 of a video encoding method according to the present invention. As shown in FIG. 3, the method includes the following steps:
  • Step 100 Obtain, according to the disparity vector information of the predetermined search precision, an image block corresponding to the current macroblock and having the same size as the search precision in the reference image of the adjacent view;
  • the corresponding macroblocks that are optimal in coding efficiency of the current macroblock may be utilized in the reference image of the neighboring view of the current macroblock to be encoded.
  • the motion information is used as the motion information of the current macroblock, so it is necessary to find the corresponding macroblock in the reference image that meets the requirements.
  • the search precision includes, for example, 8 x 8 pixel precision or 16 x 16 pixel precision, etc., and initially locates the same size as the search precision in the reference image with the current adjacent view.
  • the image block that is, if the current macroblock is in accordance with the disparity vector information of 8 x 8 pixel precision, an 8 x 8 image block can be initially located in the adjacent view reference image, if the current macro block is in accordance with the disparity vector of 16 x 16 pixel precision
  • the information may initially locate a 16 ⁇ 16 image block or the like in the reference image of the adjacent view.
  • Step 101 Establish a coordinate system of a search area of a reference image according to the image block; after initially positioning an image block in a reference image adjacent to the current macroblock, according to the positioning image block in the reference image A coordinate system is established in the search area in .
  • the search area will include the above-described positioning image block.
  • the first 8 X 8 or 4 X 4 image block of the block is used as the coordinate origin of the coordinate system of the search area of the reference image; the 8 X 8 or 4 x 4 image block may also be the coordinate origin of the coordinate system of the search area of the reference image ; ; when locating an image block as a 16 x 16 image block, using that image block as The coordinate origin of the coordinate system of the search area of the reference image; from the above, according to the size of the image block found in the reference image, the coordinate origin of the coordinate system is determined differently, and of course, it is not limited to the coordinate origin.
  • the determining manner may also be that a certain image block around the image block or a macroblock in which the image block is located is used as the coordinate origin of the coordinate system of the search area in the reference image.
  • Step 102 Search, in the search area, a corresponding macroblock that optimizes coding efficiency of the current macroblock, and obtain first offset information of the corresponding macroblock in the coordinate system.
  • the corresponding macroblocks that make the current macroblock coding efficiency optimal are searched one by one, specifically Predicting the motion information for each macroblock, obtaining residual information according to motion information of the current macroblock, and then calculating bit overhead information in the case of using the MSM mode; if the macroblock has a minimum bit overhead, the macroblock is Corresponding macroblocks that optimize the current macroblock coding efficiency in the search region in the reference image; after determining the corresponding macroblock that optimizes the current macroblock encoding, obtaining the corresponding macroblock in the already established coordinate system
  • the first coordinate information includes first offset information of the corresponding macroblock in a horizontal and vertical direction with respect to a coordinate system origin.
  • Step 103 Encode the first offset information.
  • the motion information of the corresponding macroblock that makes the current macroblock coding efficiency optimal in the reference image adjacent to the current macroblock is applied as the motion information of the current macroblock, and because the current macroblock is in the current macroblock.
  • the motion information of all the macroblocks in the reference image adjacent to the current macroblock has been encoded, that is, the motion information of the corresponding macroblock as the current macroblock motion compensation has been encoded, so Now, as long as the offset information of the corresponding macroblock in the reference image relative to the coordinate origin is encoded, the decoder is informed, and the decoder can accurately locate the corresponding macroblock according to the offset information, and the motion of the corresponding macroblock that has been decoded. The information is taken out as motion information of the current macroblock.
  • the first offset information for identifying the offset is to be encoded, including first determining the current macro.
  • the peripheral block of the block is the offset information of the corresponding macroblock in the reference image, such as the second offset information of the corresponding macroblock in the reference image of the macroblock where the left block of the current macroblock is located, and the current macro.
  • the third offset information of the corresponding macroblock of the macroblock in the reference image is located in the reference image, and then constructs the encoding context according to the obtained second and third offset information, and finally the current encoding context based on the constructed encoding context
  • the macroblock performs encoding processing on the first offset information of the corresponding macroblock in the reference image, specifically, after constructing the encoding context according to the obtained second and third offset information, the first offset information is to be used.
  • the horizontal offset and the vertical offset are binarized according to the truncated unary code or the exponential Golomb code, and the binary bit stream information is obtained, and the binary bit stream including the binarized information is sent to the arithmetic coder. Performing arithmetic coding according to the encoding context information; or directly applying a truncated unary code or an exponential Golomb code to encode each component of the first offset information into a code stream.
  • Corresponding macroblocks can be searched in the forward-view reference image or in the backward-view reference image.
  • the decoding end can accurately locate the corresponding macroblock position, so after encoding the first offset information, the marker symbol information for marking the forward-looking view is also performed.
  • Encoding specifically, XORing the tag symbol of the current macroblock with the tag symbol of one or more macroblocks in the periphery thereof, and then establishing a context model according to the tag symbol of one or more macroblocks in the periphery thereof,
  • the processed tag symbol information is encoded.
  • FIG. 4 is a schematic diagram of coordinate origin selection and offset encoding processing of a search area in the second embodiment of the video encoding method of the present invention.
  • the current macroblock MB is in a neighboring view according to a disparity vector of 8 ⁇ 8 pixel precision. Initially locate a block in the reference image (the arrow points to the display), and use the first 8 ⁇ 8 image block (shown by the black square in the figure) of the macro block where the 8 x 8 image block is located as the coordinate origin, and the search area in the shaded portion. Establish a coordinate system coordinate system.
  • FIG. 5 is a current macroblock in the second embodiment of the video encoding method of the present invention.
  • the left block A around the current macro block and the upper block B on the upper block are 4 x 4 image blocks; the two coordinate components of the current macro block are "horOffset""and"verOffset" are encoded; since the selected coordinate origin is at the center of the search area, there is a fixed upper limit for the absolute values of the horizontal and vertical components of the corresponding macroblock offset.
  • the level of the offset The absolute values of the component and vertical components will not exceed "4".
  • "horOffset” and “verOffset” are respectively binarized according to the truncated unary code, and then the binarized code stream is binarized.
  • the arithmetic coder is sent to the arithmetic coder for arithmetic coding according to the constructed context model.
  • the pseudo code for the above encoding process is as follows: xWriteOffsetComponent( Short sOffsetComp, the absolute value of the Ulnt A and B offset components And uiAbsSum, UInt context index uiCtx)
  • uiLocalCtx + ( uiAbsSum > 5) ? 3 : 2;
  • the 8 x 8 image block initially positioned in the reference image of the adjacent view may be used as the coordinate origin of the coordinate system, although the determination of the coordinate origin is different, but the subsequent The encoding process of the offset information of the corresponding macroblock of the current macroblock is the same.
  • FIG. 6 is a schematic diagram of coordinate origin selection and offset encoding processing of a search area in a third embodiment of a video encoding method according to the present invention.
  • the current macroblock MB is referenced to an adjacent view according to a disparity vector of 16 ⁇ 16 pixel precision.
  • a 16x16 block is initially positioned in the image, and the 16x16 block macroblock (shown by the black square in the figure) is used as the coordinate origin, and a two-dimensional coordinate system is established in the search area of the shaded portion.
  • AhorOffset and AverOffset the prediction residuals AhorOffset and AverOffset are obtained; and the encoding context is selected by using the offset information of the left block A and the upper block B, respectively, and AhorOffset and AverOffset are binarized according to the exponential Columbus code, respectively.
  • the binarized code stream is sent to an arithmetic coder for arithmetic coding.
  • the encoding processing method of the label symbol of the current coded macroblock in this embodiment is the same as that of the foregoing embodiment, and details are not described herein again.
  • the vertical and horizontal coordinate position information of each block in the search area is established by selecting the coordinate origin of the appropriate search area; and the information of the current block of the currently coded macro block is used as the current macro block in the phase.
  • the context of the positional offset information of the corresponding macroblock in the reference picture of the neighboring picture encodes the offset of the current macroblock, which improves the coding efficiency.
  • FIG. 7 is a flowchart of an embodiment of a video decoding method according to the present invention. As shown in FIG. 7, the method includes the following steps:
  • Step 200 Parse the received code stream information to obtain a reference picture adjacent to the current macro block.
  • the first offset information of the macroblock corresponding to the current macroblock In the image, the first offset information of the macroblock corresponding to the current macroblock;
  • the decoding end parses the information included therein to obtain offset information of the corresponding macroblock in the reference picture of the adjacent view of the macroblock to be decoded, where the corresponding macro block is A macroblock in the reference image that makes the current macroblock efficient in encoding.
  • the process of specifically obtaining the first offset information may be: first determining second offset information and a third offset of the corresponding macroblock of the macroblock in which the left block and the upper block of the current macroblock are located in the reference image.
  • the decoder that truncates the unary code or the exponential Golomb code can be used to solve the offset information of the corresponding macroblock corresponding to the current macroblock, that is, in the horizontal and vertical directions. The offset on the top.
  • the process of parsing the first offset information further includes: first determining second offset information and third offset information of the corresponding macroblock of the macroblock in which the left block and the upper block of the current macroblock are located in the reference image. And obtaining the decoding context information according to the obtained second offset information and the third offset information, and then parsing the prediction residual information of the corresponding macroblock according to the decoding context information, in the process by using the truncated unary code or the index Columbus An arithmetic decoder of the code, each bit of the prediction residual information is solved according to the decoding context information, and finally the prediction residual information of the corresponding macroblock in the reference image of the currently decoded macroblock is obtained; and the second offset information is further obtained.
  • the amount information that is, its offset in the horizontal and vertical directions.
  • Step 201 Obtain an image block corresponding to the current macroblock in the reference image of the adjacent view according to the disparity vector information
  • the origin of the coordinate is determined, that is, the obtained offset is the offset from which block.
  • Establishing a coordinate system in the search area in the reference image is consistent with the process of establishing a coordinate system in the foregoing encoding method, that is, disparity vector information according to a predetermined search precision, An image block corresponding to the current macroblock and having the same size as the search precision is obtained in the reference image of the adjacent view, and a coordinate system of the search region of the reference image in which the image is located is established according to the image block.
  • the selection rule of the origin coordinates, the encoding end and the decoding end are pre-agreed and consistent, and the coordinate system established by the decoding end according to the image block is exactly the same as the coordinate system established by the encoding end according to the image block.
  • Step 202 Obtain coordinate information of a macroblock corresponding to the current macroblock according to the first offset information in a coordinate system of a reference image search area established according to the image block; after the coordinate system is established And determining, according to the origin coordinates and the first offset information, coordinate information of the corresponding macroblock in the coordinate system, and determining a specific position of the corresponding macroblock corresponding to the currently decoded macroblock in the reference image.
  • Step 203 Obtain motion information of a macroblock corresponding to the current macroblock according to the coordinate information, and apply the motion information to perform motion compensation.
  • the motion information of the corresponding macroblock can be taken out as the currently decoded macroblock in the decoding information of the reference image.
  • Motion information used for motion compensation of the current macroblock.
  • the process of parsing the marker symbol information for marking the forward and backward view is further included, specifically according to the current macroblock. Defining a contextual model of the one or more macroblocks of the surrounding area, and parsing the identification information of the marking symbol, the identification information of the marking symbol is one or more of the marking symbol of the current macroblock and its periphery The result information of the XOR processing of the mark symbol of the macro block; after the identification information of the mark symbol is parsed, the analysis result is XORed to obtain the mark symbol information for marking the forward and backward view.
  • the location information of the corresponding macroblock in the coordinate system is obtained, and then the motion information of the corresponding macroblock is applied as the current macroblock. Motion information improves decoding efficiency.
  • An embodiment of the present invention further provides a video encoding method, including the following steps: Step 300: XOR the mark symbol of the current macroblock for marking the forward and backward view with the mark symbol of one or more macro blocks around the same;
  • the encoding end determines, by using the foregoing or existing judgment conditions, the corresponding macroblock in the forward-view or backward-view reference image of the current macroblock, and applies the motion information of the selected corresponding macroblock as the motion information of the current macroblock.
  • the marker symbol can identify the forward-view or backward-view reference image, and the encoding end selects the marker symbol of the reference image to be XORed with the marker symbol of one or more macroblocks in the vicinity, waiting for coding.
  • Step 301 Establish a context model according to the mark symbols of the one or more macroblocks in the periphery, and apply the context model to encode the XOR-processed marker symbol information.
  • the context model is established by using the mark symbols of one or more macroblocks around the current macroblock, and the selected peripheral macroblock is the same as the macroblock selected in the above step, and the context model is established for context adaptive arithmetic coding.
  • the computer readable storage medium when executed, executes the steps including the foregoing method embodiments; and the foregoing storage medium includes: a medium that can store program codes, such as a ROM, a RAM, a magnetic disk, or an optical disk.
  • FIG. 8 is a schematic structural diagram of Embodiment 1 of a video encoding apparatus according to the present invention.
  • the apparatus includes a first module 11, a second module 12, and a third module 13, wherein the first module 11 is configured to perform predetermined search accuracy.
  • the disparity vector information, the image block corresponding to the current macroblock and having the same size as the search precision is obtained in the reference image of the adjacent view;
  • the second module 12 is configured to be in the coordinate system established according to the image block, Obtaining first offset information of the corresponding macroblock that makes the current macroblock coding efficiency optimal;
  • the third module 13 is configured to encode the first offset information.
  • the first module 11 in the video encoding apparatus initially specifies an image block in the reference image according to the search precision disparity vector information, the size of the image block being the same as the search precision; and then the second module 12 is based on the image block.
  • a two-dimensional coordinate system is established in the search area in the reference image, and all macroblocks in the reference image through the coordinate system have position information; when corresponding macroblocks that make the coding efficiency of the current macroblock optimal are found according to a certain search order Obtaining first offset information of the corresponding macroblock, that is, offset information relative to the coordinate origin; the third module 13 encodes the first offset information, and further, the third module 13 includes the first sub- The module 131, the second sub-module 132, and the third sub-module 133, when the first sub-module 131 determines that the peripheral block of the current macroblock, for example, the macroblock in which the left block and the upper block are respectively located, is in the second of the corresponding macroblock in the reference image.
  • the second sub-module 132 is configured to obtain the encoding according to the second offset information and the third offset information.
  • the third sub-module 133 for applying a first encoding context information offset information encoding process.
  • the fourth module 14 is further configured to encode the marker symbol information for marking the forward and backward view.
  • the fourth module 14 includes the eighth submodule 141 and the ninth submodule. 142, wherein the eighth sub-module 141 performs an exclusive-OR processing of the mark symbol of the current macroblock for marking the forward-backward view marker symbol with one or more macroblocks of the periphery thereof, the ninth sub-module 142 according to the periphery
  • the tag symbols of one or more macroblocks establish a context model that encodes the XOR-processed tag symbol information.
  • FIG. 9 is a schematic structural diagram of Embodiment 2 of a video encoding apparatus according to the present invention, and FIG.
  • the video encoding device provided by the embodiment is different from the video encoding device embodiment 1 in that the third module 13 includes a fourth submodule 134, a fifth submodule 135, a sixth submodule 136, and a seventh submodule 137.
  • the fifth sub-module 135 averages the corresponding components of the second offset information and the third offset information, and applies the average processing result to predict the first offset information to obtain prediction residual information;
  • the seventh sub-module 137 applies the encoding context information to perform encoding processing on the prediction residual information.
  • the vertical and horizontal coordinate position information of each block in the search area is established by selecting the coordinate origin of the appropriate search area; using the information of the current block of the currently coded macro block as the coded current macro block in the adjacent view
  • the context of the position offset information of the corresponding macroblock in the reference picture encodes the offset of the current macroblock, which improves the coding efficiency.
  • FIG. 10 is a schematic structural diagram of Embodiment 1 of a video decoding apparatus according to the present invention.
  • the apparatus includes: a fifth module 21, a sixth module 22, a seventh module 23, and an eighth module 24, wherein the fifth module 21 is used. And parsing the received code stream information to obtain first offset information of a macroblock corresponding to the current macroblock in a reference image adjacent to the current macroblock; and the sixth module 22 is configured to use the disparity vector information.
  • the seventh module 23 is configured to, according to the coordinate system of the reference image search area established according to the image block, according to the first offset Obtaining coordinate information of the macroblock corresponding to the current macroblock; the eighth module 24 is configured to obtain motion information of the macroblock corresponding to the current macroblock according to the coordinate information, and apply the motion information to perform motion compensation .
  • the fifth module 21 in the device parses the offset information of the corresponding macroblock in the reference image of the currently decoded macroblock, and the seventh module 23 is configured according to the image found in the sixth module 22.
  • the block establishes a two-dimensional coordinate system within the search area of the reference image, The coordinate information to the corresponding macroblock; the eighth module 24 extracts the motion information of the corresponding macroblock in the motion information of all the macroblocks of the already decoded reference image, and performs motion compensation as the motion information of the current macroblock.
  • the fifth module 21 includes a tenth submodule 211, an eleventh submodule 212, and a twelfth submodule 213, wherein the tenth submodule 211 is configured to determine a peripheral block of the current macroblock, such as a left block and an upper block. a second offset information and a third offset information of the corresponding macroblock in the reference image; the eleventh submodule 212 is configured to obtain the decoding according to the second offset information and the third offset information.
  • Context information The twelfth sub-module 213 is configured to apply the decoding context information to parse the first offset information.
  • the apparatus further includes a ninth module 25 for parsing the marker symbol information for marking the forward and backward view. After receiving the code stream information, the application ninth module 25 parses the marker symbol information in the code stream information to determine Which of the forward-looking reference images the corresponding macroblock of the currently decoded macroblock is located in.
  • the fifth module 21 includes a thirteenth submodule 214 and a fourteenth submodule 215. a fifteenth sub-module 216 and a sixteenth sub-module 217, wherein the thirteenth sub-module 214 is configured to determine that a peripheral block of the current macroblock, for example, a macroblock in which the left block and the upper block are respectively located, corresponds to the reference image a second offset information of the macroblock and a third offset information; the fourteenth submodule 215 is configured to obtain decoding context information according to the second offset information and the third offset information; The module 216 is configured to parse the prediction residual information of the corresponding macroblock according to the decoding context information. The sixteenth submodule 217 is configured to use the corresponding component of the second offset information and the third offset information. Performing an averaging process, and obtaining first offset information of the corresponding macroblock according to the processing result and the
  • the position information of the corresponding macroblock in the coordinate system is obtained, and then the motion information of the corresponding macroblock is applied as the current macro.
  • the motion information of the block improves the decoding efficiency.
  • the system includes a video encoding apparatus 1 and a video decoding apparatus 2, wherein the video encoding apparatus 1 includes a first module 11, a second module 12, and a a third module 13 , wherein the first module 11 is configured to obtain an image block corresponding to the current macroblock and having the same size as the search precision in the reference image of the adjacent view according to the disparity vector information of the predetermined search precision;
  • the second module 12 is configured to obtain, in a coordinate system established according to the image block, first offset information of a corresponding macroblock that makes the current macroblock encoding efficiency optimal;
  • the third module 13 is configured to An offset information is encoded.
  • the first module 11 in the video encoding apparatus 1 initially specifies an image block in the reference image according to the search precision disparity vector information, the size of the image block being the same as the search precision; and then the second module 12 is based on the image block.
  • the first offset information of the corresponding macroblock that is, the offset information relative to the coordinate origin is obtained;
  • the third module 13 encodes the first offset information, and further, the third module 13 includes the first The sub-module 131, the second sub-module 132, and the third sub-module 133, when the first sub-module 131 determines that the peripheral block of the current macroblock, for example, the macroblock in which the left block and the upper block are respectively located, the corresponding macroblock in the reference image
  • the video encoding apparatus 1 further includes a fourth module 14 for encoding the marker symbol information for marking the forward and backward views.
  • the fourth module 14 includes the eighth submodule 141 and The ninth sub-module 142, wherein the eighth sub-module 141 performs XOR processing on the mark symbol of the current macroblock for marking the forward-backward view mark symbol with one or more macroblocks of the periphery thereof, the ninth sub-module 142
  • the XOR-processed tag symbol information is encoded according to the tagging symbols of the surrounding one or more macroblocks.
  • the video decoding device 2 includes a fifth module 21, a sixth module 22, a seventh module 23, and an eighth module 24, wherein the fifth module 21 is configured to parse the received code stream information to obtain a reference adjacent to the current macroblock.
  • the first offset information of the macroblock corresponding to the current macroblock In the image, the first offset information of the macroblock corresponding to the current macroblock; the sixth module 22 is configured to obtain an image block corresponding to the current macroblock in the reference image of the adjacent view according to the disparity vector information;
  • the seventh module 23 is configured to obtain coordinate information of a macroblock corresponding to the current macroblock according to the first offset information in a coordinate system of a reference image search area established according to the image block; 24, configured to obtain motion information of a macroblock corresponding to the current macroblock according to the coordinate information, and apply the motion information to perform motion compensation.
  • the fifth module 21 in the video decoding device 2 parses the offset information of the corresponding macroblock in the reference image of the currently decoded macroblock, and the seventh module 23 is configured according to the sixth module 22.
  • the found image block establishes a two-dimensional coordinate system within the search area of the reference image to obtain coordinate information of the corresponding macroblock;
  • the eighth module 24 extracts the motion of the corresponding macroblock from the motion information of all macroblocks of the already decoded reference image. Information, motion compensation as motion information of the current macroblock.
  • the fifth module 21 includes a tenth submodule 211, an eleventh submodule 212, and a twelfth submodule 213, wherein the tenth submodule 211 is configured to determine a peripheral block of the current macroblock, such as a left block and an upper block. a second offset information and a third offset information of the corresponding macroblock in the reference image; the eleventh submodule 212 is configured to obtain the decoding according to the second offset information and the third offset information.
  • Context information The twelfth sub-module 213 is configured to apply the decoding context information to parse the first offset information.
  • the video decoding device 2 further includes a ninth module 25 for parsing the marker symbol information for marking the forward and backward view. After receiving the code stream information, it first determines whether there is the encoded information of the marker symbol. If the person exists, the application is applied. The ninth module 25 parses the tag symbol information to determine which of the view reference images the corresponding macroblock of the currently decoded macroblock is located in.
  • FIG. 13 is a schematic structural diagram of Embodiment 2 of a video processing system according to the present invention.
  • the system includes a video encoding apparatus 1 and a video decoding apparatus 2, where the video encoding apparatus 1 includes a module 1 1 , a second module 12 , and a third module 13 , wherein the first module 11 is configured to obtain, according to a predetermined search accuracy, disparity vector information, corresponding to the current macro block in the reference image of the adjacent view, and And the second module 12 is configured to obtain first offset information of a corresponding macroblock that optimizes encoding efficiency of the current macroblock in a coordinate system established according to the image block.
  • the third module 13 is configured to encode the first offset information.
  • the first module 11 in the video encoding apparatus 1 initially specifies an image block in the reference image according to the search precision disparity vector information, the size of the image block being the same as the search precision; and then the second module is based on the image block.
  • a two-dimensional coordinate system is established in the search area in the reference image, and all macroblocks in the reference image through the coordinate system have position information; when corresponding macroblocks that make the coding efficiency of the current macroblock optimal are found according to a certain search order Obtaining first offset information of the corresponding macroblock, that is, offset information relative to the coordinate origin; the third module 13 encodes the first offset information, and further the third module 3 includes the fourth submodule 134.
  • the fifth submodule 135 sets the second offset information and the third offset information The corresponding component of the interest is averaged, and the first offset information is predicted by using the average processing result to obtain prediction residual information; and the sixth sub-module 136 obtains the second offset information and the third offset information.
  • the seventh sub-module 137 applies the encoding context information to encode the prediction residual information.
  • the video encoding apparatus 1 further includes a fourth module 14 for encoding the tag symbol information for marking the forward and backward view.
  • the fourth module 14 includes the eighth submodule 141 and The ninth sub-module 142, wherein the eighth sub-module 141 performs XOR processing on the mark symbol of the current macroblock for marking the forward-backward view mark symbol with one or more macroblocks of the periphery thereof, the ninth sub-module 142
  • the XOR-processed tag symbol information is encoded according to the tagging symbols of the surrounding one or more macroblocks.
  • the video decoding device 2 includes a fifth module 21, a sixth module 22, a seventh module 23, and an eighth a module 24, wherein the fifth module 21 is configured to parse the received code stream information, to obtain a first offset information of a macroblock corresponding to the current macroblock in a reference image adjacent to the current macroblock;
  • the sixth module 22 is configured to obtain an image block corresponding to the current macroblock in the reference image of the adjacent view according to the disparity vector information;
  • the seventh module 23 is configured to search the coordinate system of the reference image according to the image block.
  • the eighth module 24 is configured to obtain motion information of a macroblock corresponding to the current macroblock according to the coordinate information. Applying the motion information to perform motion compensation.
  • the fifth module 21 in the device parses the offset information of the corresponding macroblock in the reference image of the currently decoded macroblock, and the seventh module 23 is configured according to the image found in the sixth module 22.
  • the block establishes a two-dimensional coordinate system within a search area of the reference image to obtain coordinate information of the corresponding macroblock;
  • the eighth module 24 extracts motion information of the corresponding macroblock from the motion information of all the macroblocks of the already decoded reference image, as The motion information of the current macroblock is motion compensated.
  • the fifth module 21 includes a thirteenth submodule 214, a fourteenth submodule 215, a fifteenth submodule 216, and a sixteenth submodule 217, wherein the thirteenth submodule 214 is configured to determine the current macro a peripheral block of the block, for example, a second offset information and a third offset information of a corresponding macroblock of the macroblock in which the macroblock and the upper block are located in the reference image; the fourteenth submodule 215 is configured to The second offset information and the third offset information obtain decoding context information; the fifteenth submodule 216 is configured to parse the prediction residual information of the corresponding macroblock according to the decoding context information; The 217 is configured to perform averaging processing on the corresponding components of the second offset information and the third offset information, and obtain first offset information of the corresponding macroblock according to the processing result and the prediction residual information.
  • the thirteenth submodule 214 is configured to determine the current macro a peripheral block of the block, for example, a second offset information and a
  • the apparatus further includes a ninth module 25 for parsing the marker symbol information for marking the forward and backward view. After receiving the code stream information, first determining whether there is the coded information of the marker symbol, and if the person exists, applying the ninth The module 25 parses the marker symbol information to determine which of the forward reference images the corresponding macroblock of the currently decoded macroblock is located in.
  • the video encoding apparatus establishes the vertical and horizontal coordinate position information of each block in the search area by selecting the coordinate origin of the appropriate search area; using the information of the peripheral block of the current coded macro block as the current code.
  • the macroblock encodes the offset of the current macroblock in the context of the position offset information of the corresponding macroblock in the adjacent reference picture, and improves the coding efficiency;
  • the video decoding apparatus parses the offset of the corresponding macroblock of the current macroblock by analyzing The quantity information obtains the position information of the corresponding macroblock in the coordinate system, and then applies the motion information of the corresponding macroblock as the motion information of the current macroblock, thereby improving the decoding efficiency.

Landscapes

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

Description

视频编码、 解码方法及装置和视频处理*** 技术领域
本发明实施例涉及视频技术领域, 尤其涉及一种视频编码方法及装 置, 视频解码方法及装置和视频处理***。 背景技术
随着多媒体通信技术的发展, 传统的固定视点视觉以及 2D平面视觉 已经不能满足人们对视频播放的高要求, 在娱乐、 教育、 观光和外科医学 等诸多应用领域均出现了对于自由视点视频和 3D视频的需求, 例如能够 由观看者选择观看视角的自由视点电视 ( Free view-point Television ; 以 下简称: FTV ) , 以及为处于不同位置的观看者提供不同视角视频的立体 视觉电视( 3 Dimensional Television; 以下简称: 3DTV )等。 目前 ITU和 MPEG的联合视频工作组正在制定的与 H.264/AVC兼容的联合多视点视 频编码技术标准中, 联合多视点视频编码模型 (Joint Multiview Video Model ; 以下简称: JMVM ) 釆用了视点间预测的运动信息跳跃模式 ( Motion Skip Mode ; 以下简称: MSM ) , 该技术利用相邻视点视图中 运动的高度相似性,将相邻视点视图中的运动信息用于当前视点视图的编 码, 可以节省编码图像中某些宏块运动信息所需的比特资源, 从而提高多 视点视频编码 ( Multi- view Video Coding ; 以下简称: MVC )压缩效率。
MSM技术主要包括计算全局视差矢量信息( Global Disparity Vector ; 以下简称: GDV )和计算参考图像中对应宏块的运动信息两个步骤。 如图 1所示, 两侧的上下两个方块表示相邻视图中的锚定帧( Anchor Picture ) , 在锚定帧 ImgA和锚定帧 ImgB之间可以有多个非锚定帧, 图 1中只表示了 一个非锚定帧 Imgeur,通过公式^ = GD 可得到非锚定帧 Imgeur的全局 视差信息 GDVcur; 在得到当前编码图像 Imgcur的 GDVcur信息后, 可以根 据该 GDVCM信息确定非锚定帧 ImgCM中每一个宏块在视点间参考视图像 中的对应宏块, 例如图中的宏块 MBCM在视点间参考视图像中的对应宏块 为 MBCOT, 并应用宏块 MBCOT的运动信息作为宏块 MBcur的运动信息, 用于 运动补偿, 在视图内找到对应参考帧的对应宏块作预测, 得到残差数据, 最后计算得到使用 MSM模式的开销 RDCostMBcM, MSM, 若 MSM模式开销 小于其它宏块模式开销, 则 MSM被选定为该宏块最终模式。
上述方法中, 通过 GDVcur信息确定对应宏块可能并不是使得当前宏 块编码效率最优的对应宏块, 为找到使当前宏块编码效率最优的对应宏 块, 可以在参考图像中预先设定的搜索范围内, 寻找使当前宏块编码效率 最优的对应宏块的运动信息, 进而得到当前宏块的运动信息。 具体如图 2 所示, 该方法中利用索引标识对搜索范围内的每个块进行查找, 索引号分 别是 0、 1、 2、 3等。 在对当前宏块 MB进行编码时, 如果在邻近视点中 的搜索范围内能够找到使编码效率最优的对应宏块 MB,,假设最优的是索 引号为 5的宏块, 则在编码当前宏块 MB时, 同时编码该宏块 MB,的索引 号 "5" 。
上述方法中, 由于要对查找到的对应宏块的索引信息进行编码, 因此 存在信息冗余; 而且由于搜索区域是二维的, 而该方法中的索引号编码方 法是一维化的位置偏移信息,没有揭示位置偏移信息在水平方向和垂直方 向各自的统计特性, 影响了编码效率。
而且, 现有技术中应用前向视或后向视的参考图像中 GDV信息指向 的对应宏块的运动信息作为当前编码宏块的运动信息, 用于当前编码宏块 的运动补偿, 但由于前向视或后向视的参考图像中对应宏块的差异, 使得 编码效率低。 发明内容
本发明实施例提供一种视频编码方法及装置, 一种视频解码方法及装 置和视频处理***, 用以解决现有技术中编码效率低的缺陷, 实现视频图 像的高效率编码。
本发明实施例提供一种视频编码方法, 包括:
根据视差矢量信息, 在相邻视的参考图像中获得与当前宏块对应的图 像块;
根据所述图像块建立该图像块所在参考图像搜索区域的坐标系; 在所述搜索区域内查找使所述当前宏块编码效率最优的对应宏块, 并 获得所述对应宏块在所述坐标系中的第一偏移量信息;
对所述第一偏移量信息进行编码。
本发明实施例提供一种视频解码方法, 包括:
解析接收到的码流信息, 得到与当前宏块相邻视的参考图像中, 与所 述当前宏块对应的宏块的第一偏移量信息;
根据视差矢量信息, 在相邻视的参考图像中获得与当前宏块对应的图 像块;
在根据所述图像块建立的参考图像搜索区域的坐标系中, 根据所述第 一偏移量信息获得与所述当前宏块对应的宏块的坐标信息;
根据所述坐标信息获得与所述当前宏块对应的宏块的运动信息, 应用 所述运动信息进行运动补偿。
本发明实施例提供一种视频编码装置, 包括:
第一模块, 用于根据预定的搜索精度的视差矢量信息, 在相邻视的参 考图像中获得与当前宏块对应的、 与所述搜索精度大小相同的图像块; 第二模块, 用于在根据所述图像块建立的坐标系中, 获得使得所述当 前宏块编码效率最优的对应宏块的第一偏移量信息;
第三模块, 用于对所述第一偏移量信息进行编码。
本发明实施例提供一种视频解码装置, 包括:
第五模块, 用于解析接收到的码流信息, 得到与当前宏块相邻视的参 考图像中, 与所述当前宏块对应的宏块的第一偏移量信息; 第六模块, 用于根据视差矢量信息, 在相邻视的参考图像中获得与当 前宏块对应的图像块;
第七模块, 用于在根据所述图像块建立的参考图像搜索区域的坐标系 中, 根据所述第一偏移量信息获得与所述当前宏块对应的宏块的坐标信 息;
第八模块, 用于根据所述坐标信息获得与所述当前宏块对应的宏块的 运动信息, 应用所述运动信息进行运动补偿。
本发明实施例提供一种视频处理***, 包括视频编码装置和视频解码 装置, 所述视频编码装置包括:
第一模块, 用于根据预定的搜索精度的视差矢量信息, 在相邻视的参 考图像中获得与当前宏块对应的、 与所述搜索精度大小相同的图像块; 第二模块, 用于在根据所述图像块建立的坐标系中, 获得使得所述当 前宏块编码效率最优的对应宏块的第一偏移量信息;
第三模块, 用于对所述第一偏移量信息进行编码;
所述视频解码装置包括:
第五模块, 用于解析接收到的码流信息, 得到与当前宏块相邻视的参 考图像中, 与所述当前宏块对应的宏块的第一偏移量信息;
第六模块, 用于根据视差矢量信息, 在相邻视的参考图像中获得与当 前宏块对应的图像块;
第七模块, 用于在根据所述图像块建立的参考图像搜索区域的坐标系 中, 根据所述第一偏移量信息获得与所述当前宏块对应的宏块的坐标信 息;
第八模块, 用于根据所述坐标信息获得与所述当前宏块对应的宏块的 运动信息, 应用所述运动信息进行运动补偿。
本发明实施例提供一种视频编码方法, 包括: 将当前宏块的用于标记前后向视的标记符号与其周边的一个或多个 宏块的标记符号进行异或处理;
根据所述周边的一个或多个宏块的标记符号建立上下文模型, 应用所 述上下文模型对所述经过异或处理的标记符号信息进行编码。
本发明实施例提供的视频编码方法及装置, 视频解码方法及装置和视 频处理***, 通过选择适当的搜索区域的坐标原点, 建立搜索区域内每个 块的纵横坐标位置信息; 同时利用当前编码宏块周边块的信息, 作为编码 当前宏块在相邻视的参考图像中对应宏块的位置偏移信息的上下文, 编码 当前宏块的偏移量, 提高了编码效率。 附图说明
图 1为现有 GDV推导编码处理示意图;
图 2为现有搜索区域范围内的位置信息编码处理示意图;
图 3为本发明视频编码方法实施例一流程图;
图 4为本发明视频编码方法实施例二中搜索区域的坐标原点选择和偏 移量编码处理示意图;
图 5为本发明视频编码方法实施例二中当前宏块的对应宏块的偏移坐 标编码示意图;
图 6为本发明视频编码方法实施例三中搜索区域的坐标原点选择和偏 移量编码处理示意图;
图 7为本发明视频解码方法实施例流程图;
图 8为本发明视频编码装置实施例一结构示意图;
图 9为本发明视频编码装置实施例二结构示意图;
图 10为本发明视频解码装置实施例一结构示意图;
图 11为本发明视频解码装置实施例二结构示意图;
图 12为本发明视频处理***实施例一结构示意图; 图 13为本发明视频处理***实施例二结构示意图。 具体实施方式
下面结合附图和具体实施例进一步说明本发明实施例的技术方案。 图 3为本发明视频编码方法实施例一流程图, 如图 3所示, 该方法包 括如下步骤:
步骤 100 , 根据预定的搜索精度的视差矢量信息, 在相邻视的参考图 像中获得与当前宏块对应的、 与所述搜索精度大小相同的图像块;
在 MSM模式中, 由于相邻视点视图中运动的高度相似性, 可以利用 与要进行编码的当前宏块的相邻视的参考图像中, 使所述当前宏块编码效 率最优的对应宏块的运动信息作为当前宏块的运动信息, 因此就要在参考 图像中找到符合要求的对应宏块。 首先根据预先设定好的搜索精度的视差 矢量信息, 搜索精度包括例如 8 x 8像素精度或 16 x 16像素精度等, 在与 当前相邻视的参考图像中初始定位一个与搜索精度大小相同的图像块, 即 若当前宏块按照 8 x 8像素精度的视差矢量信息可以在相邻视的参考图像 中初始定位一个 8 x 8的图像块, 若当前宏块按照 16 x 16像素精度的视差 矢量信息可以在相邻视的参考图像中初始定位一个 16 X 16的图像块等。
步骤 101 ,根据所述图像块建立其所在参考图像的搜索区域的坐标系; 当在与当前宏块相邻视的参考图像中初始定位到一个图像块之后, 根 据定位图像块在所述参考图像中的搜索区域内建立坐标系。 其中参考图像 中的搜索区域的范围是预先设定好的, 该搜索区域将包括上述的定位图像 块。 根据定位图像块在参考图像中的搜索区域内建立一个二维坐标系, 具 体可以为, 当定位图像块为 8 x 8或 4 x 4图像块时, 以该图像块或以该图 像块所在宏块的首个 8 X 8或 4 X 4图像块作为参考图像的搜索区域的坐标 系的坐标原点; 也可以该 8 X 8或 4 x 4图像块为参考图像的搜索区域的坐 标系的坐标原点; ; 当定位图像块为 16 x 16图像块时, 以该图像块作为 所述参考图像的搜索区域的坐标系的坐标原点; 由上可知, 根据在参考图 像中找到的图像块的大小不同, 坐标系的坐标原点的确定也不同, 当然也 不局限于上述坐标原点的确定方式, 还可以是以定位图像块周边的某个图 像块或其所在宏块作为参考图像内搜索区域的坐标系的坐标原点。
步骤 102 , 在所述搜索区域内查找使所述当前宏块编码效率最优的对 应宏块, 并获得所述对应宏块在所述坐标系中的第一偏移量信息;
在确定完坐标系的原点坐标后, 在参考图像中搜索区域范围内, 按照 从左向右, 从上向下的搜索方式, 逐个查找使得当前宏块编码效率最优的 对应宏块, 具体为对每一宏块进行欲动信息的预测, 根据当前宏块的运动 信息得到残差信息, 然后计算在使用 MSM模式情况下的比特开销信息; 若一宏块的比特开销最小则将该宏块作为在参考图像中搜索区域范围内 使得当前宏块编码效率最优的对应宏块; 确定使当前宏块编码最优的对应 宏块后, 获得该对应宏块在已经建立好的坐标系中的第一坐标信息, 所述 的第一坐标信息包括对应宏块相对于坐标系原点在水平和垂直方向上的 第一偏移信息。
步骤 103 , 对所述第一偏移量信息进行编码。
因为在 MSM模式中, 是应用与当前宏块相邻视的参考图像中使得该 当前宏块编码效率最优的对应宏块的运动信息作为当前宏块的运动信息, 而且因为在对当前宏块进行编码处理之前, 与当前宏块相邻视的参考图像 中所有宏块的运动信息都已经经过编码处理了, 即作为当前宏块运动补偿 的对应宏块的运动信息已经经过编码处理了, 因此现在只要将对应宏块在 参考图像中相对于坐标原点的偏移信息进行编码, 告知解码器, 解码器便 可以按照该偏移信息准确定位对应宏块, 并将已经解码的对应宏块的运动 信息取出作为当前宏块的运动信息。
在获得当前宏块在参考图像中的对应宏块的第一偏移量信息后, 要对 该用于标识偏移量的第一偏移量信息进行编码处理, 包括首先确定当前宏 块的周边块如所在宏块在参考图像中对应的宏块的偏移量信息, 例如当前 宏块的左边块所在宏块在参考图像中对应的宏块的第二偏移量信息和当 前宏块的上边块所在宏块在参考图像中对应的宏块的第三偏移量信息, 然 后根据得到的第二和第三偏移量信息构造编码上下文, 最后再根据构造好 的编码上下文对当前宏块在参考图像中的对应宏块的第一偏移量信息进 行编码处理, 具体为, 在根据得到的第二和第三偏移量信息构造编码上下 文后, 要将第一偏移量信息中的水平偏移量和垂直偏移量按照截断一元码 或指数哥伦布码把偏移信息二值化, 得到二进制位流信息, 再将包括二值 化信息的二进制位流发送到算术编码器, 根据所述编码上下文信息进行算 术编码; 或者直接应用截断一元码或指数哥伦布码将所述第一偏移量信息 各分量编码进码流。
根据构造好的编码上下文对当前宏块在参考图像中的对应宏块的第 一偏移量信息进行编码处理, 还可以为首先确定当前宏块的左边块和上边 块各自所在宏块在参考图像中对应的宏块的第二偏移量信息和第三偏移 量信息, 然后将第二偏移量信息和第三偏移量信息的对应分量进行平均处 理, 即将第二偏移量信息和第三偏移量信息中的水平偏移分量进行平均, 获得水平方向均值; 将第二偏移量信息和第三偏移量信息中的垂直偏移分 量进行平均, 获得垂直方向均值; 再应用得到的水平偏移均值和垂直偏移 均值对第一偏移量信息对应分量进行预测, 并得到预测残差信息; 然后根 据第二偏移量信息和第三偏移量信息构造编码上下文信息, 应用该编码上 下文信息对预测残差信息进行编码处理, 具体为将得到的预测残差信息按 照截断一元码或指数哥伦布码把偏移信息二值化, 再将包括二值化信息的 码流发送到算术编码器, 根据所述编码上下文信息进行算术编码; 或者直 接应用截断一元码或指数哥伦布码将所述第一偏移量信息各分量编码进 码流。
因为在参考图像中搜索区域范围内查找使得当前宏块编码效率最优 的对应宏块, 既可以是在前向视的参考图像中查找, 也可以是在后向视的 参考图像中查找; 在对当前宏块进行编码时, 要明确告知解码端对应宏块 是位于前向视或后向视的参考图像中, 解码端才能够准确定位对应宏块位 置, 因此在对第一偏移量信息进行编码之后, 还要对用于标记前后向视的 标记符号信息进行编码, 具体为将当前宏块的标记符号与其周边的一个或 多个宏块的标记符号进行异或处理, 再根据其周边的一个或多个宏块的标 记符号建立上下文模型, 对经过异或处理的标记符号信息进行编码。 上述 码处理。
图 4为本发明视频编码方法实施例二中搜索区域的坐标原点选择和偏 移量编码处理示意图, 如图 4所示, 当前宏块 MB按照 8 x 8像素精度的 视差矢量在相邻视的参考图像中初始定位一块(箭头指向所示) , 并以该 8 x 8图像块所在宏块的首个 8 χ 8图像块(图中黑方块所示)为坐标原点, 在阴影部分的搜索区域内建立坐标系坐标系。 在搜索区域内查找使得当前 宏块编码效率最优的对应宏块, 例如该对应宏块 MB'的坐标是 (horOffset, verOffset); 图 5为本发明视频编码方法实施例二中当前宏块的对应宏块的 偏移坐标编码示意图, 如图 5所示, 以当前宏块周边的左边块 A和上边块 上边块 B为 4 x 4的图像块; 对当前宏块的两个坐标分量 "horOffset" 和 "verOffset" 进行编码; 由于选择的坐标原点在搜索区域的中心, 对应宏 块偏移量的水平分量和垂直分量的绝对值有固定的上限, 对图 5而言, 偏 移量的水平分量和垂直分量的绝对值不会超过 "4" 。 在利用左块 A和上 块 B的偏移量信息选定了编码上下文后,分别将 "horOffset"和" verOffset" 按照截断一元码把这些符号二值化, 再把经过二值化的码流发送给算术编 码器按构建好的上下文模型进行算术编码。 上述编码过程的伪代码如下: xWriteOffsetComponent( Short sOffsetComp, Ulnt A和 B偏移量分量的绝对值 和 uiAbsSum, UInt上下文索引 uiCtx )
{
//—设立上下文 - UInt uiLocalCtx = uiCtx;
if(uiAbsSum >= 3)
{
uiLocalCtx += ( uiAbsSum > 5) ? 3 : 2;
}
//--- 首符号非零否 - UInt uiSymbol = ( 0 == sOffsetComp) ? 0 : 1 ;
writeSymbol( uiSymbol, m_cOffsetCCModel.get( 0, uiLocalCtx ) );
ROTRS( 0 == uiSymbol, Err::m_nOK );
II— 非零绝对值和符号
UInt uiSign = 0;
if( 0 > sOffsetComp )
{
uiSign = 1;
sOffsetComp = -sOffsetComp;
}
按截断一元码二值化 (sOffsetComp- 1)并按一定的上下文模型进行算术编码; }
若搜索在前向视和后向视的参考图像中都要进行, 则还需要编码用于标 记前后向视的标记符号, 把当前编码宏块的标记符号 "currFlag" 和周边一个 或多个宏块的标记符号 "leftFlag" 作异或操作后, 建立上下文模型进行上下 文自适应算术编码。 其伪代码如下:
uiSymbol = currFlag异或 leftFlag;
uiCtx = (leftFlag==LIST_0) ? 0 : 1;
uiCtx += (aboveFlag==LIST_0) ? 0 : 1; writeSymbol( uiSymbol, MotionSkipListXFlagCCModel.get( 0, uiCtx ) );
在上述方法实施中, 也可以 8 χ 8像素精度的视差矢量在相邻视的参考 图像中初始定位的 8 x 8图像块, 作为坐标系的坐标原点, 虽然坐标原点的 确定不同, 但后续对当前宏块的对应宏块的偏移量信息的编码处理过程相 同。
图 6为本发明视频编码方法实施例三中搜索区域的坐标原点选择和偏 移量编码处理示意图, 如图 6所示, 当前宏块 MB按照 16x16象素精度的 视差矢量在相邻视的参考图像中初始定位一个 16x16块, 以该 16x16块的 宏块(图中黑方块所示)为坐标原点, 在阴影部分的搜索区域内建立二维 坐标系。 在搜索区域内查找使得当前宏块编码效率最优的对应宏块, 例如 找到的最优的对应宏块 MB,的坐标是 (horOffset, verOffset); 如图 5所示, 利用当前宏块左边块 A和上边块 B的偏移量的对应分量的均值预测
"horOffset" 和 "verOffset" , 得到预测残差 AhorOffset和 AverOffset; 再 利用左边块 A和上边块 B的偏移量信息选定编码上下文,分别把 AhorOffset 和 AverOffset按照指数哥伦布码进行二值化, 再把经过二值化的码流送到 算术编码器进行算术编码。 本实施例中对当前编码宏块的标记符号的编码 处理方法与上述实施例相同, 此处不再赘述。
上述各视频编码方法实施例中, 通过选择适当的搜索区域的坐标原 点, 建立搜索区域内每个块的纵横坐标位置信息; 同时利用当前编码宏块 周边块的信息, 作为编码当前宏块在相邻视的参考图像中对应宏块的位置 偏移信息的上下文, 编码当前宏块的偏移量, 提高了编码效率。
视频解码方法实施例
图 7为本发明视频解码方法实施例流程图, 如图 7所示, 该方法包括 如下步骤:
步骤 200, 解析接收到的码流信息, 得到与当前宏块相邻视的参考图 像中, 与所述当前宏块对应的宏块的第一偏移量信息;
解码端在接收到码流信息后对其中包括的信息进行解析, 得到当前所 要进行解码的宏块在相邻视的参考图像中的对应宏块的偏移量信息, 所述 对应宏块为在参考图像中使得当前宏块进行编码时效率最优的宏块。 具体 解析得到第一偏移量信息的过程可以为: 首先确定当前宏块的左边块和上 边块各自所在宏块在参考图像中对应的宏块的第二偏移量信息和第三偏 移量信息, 并根据得到的第二偏移量信息和第三偏移量信息获得解码上下 文信息, 由算术解码器按获得的解码上下文信息解出第一偏移量信息的每 个位, 得到第一偏移量信息; 在解析每一个位的过程中, 可以使用截断一 元码或指数哥伦布码的解码器解出与当前宏块对应的对应宏块的偏移量 信息, 即其在水平和垂直方向上的偏移量。
解析得到第一偏移量信息的过程还包括: 首先确定当前宏块的左边块 和上边块各自所在宏块在参考图像中对应的宏块的第二偏移量信息和第 三偏移量信息, 然后根据得到的第二偏移量信息和第三偏移量信息获得解 码上下文信息, 再根据解码上下文信息解析得到对应宏块的预测残差信 息, 此过程中由使用截断一元码或指数哥伦布码的算术解码器, 按解码上 下文信息解出预测残差信息的每个位, 最终得到当前解码宏块的在参考图 像中的对应宏块的预测残差信息; 再将第二偏移量信息和第三偏移量信息 的对应分量进行平均处理得到水平和垂直方向两个平均值后, 根据该平均 值和先前得到的预测残差信息共同获得与当前宏块对应的对应宏块的偏 移量信息, 即其在水平和垂直方向上的偏移量。
步骤 201 , 根据视差矢量信息, 在相邻视的参考图像中获得与当前宏 块对应的图像块;
在得到对应宏块的偏移量信息后, 要确定坐标原点, 即得到的偏移量 是相对于哪个块的偏移。 建立参考图像中搜索区域内的坐标系与前述编码 方法中建立坐标系的过程一致, 即根据预定的搜索精度的视差矢量信息, 在相邻视的参考图像中获得与当前宏块对应的、 与搜索精度大小相同的图 像块, 并根据该图像块建立其所在参考图像的搜索区域的坐标系。 建立坐 标系时原点坐标的选择规则, 编码端和解码端是预先约定好即保持一致 的, 解码端根据图像块建立的坐标系与编码端根据图像块建立的坐标系是 完全相同的。
步骤 202, 在根据所述图像块建立的参考图像搜索区域的坐标系中, 根据所述第一偏移量信息获得与所述当前宏块对应的宏块的坐标信息; 在建立完坐标系后, 基于原点坐标和第一偏移量信息便可确定对应宏 块在该坐标系中的坐标信息, 也就可确定与当前解码宏块对应的对应宏块 在参考图像中的具***置。
步骤 203 , 根据所述坐标信息获得与所述当前宏块对应的宏块的运动 信息, 应用所述运动信息进行运动补偿。
由于参考图像中所有宏块的运动信息均已经过解码了, 因此在确定了 对应宏块的位置后, 便可在参考图像的解码信息中取出该对应宏块的运动 信息作为当前解码宏块的运动信息, 用于当前宏块的运动补偿。
若在接收到的码流中有用于标记前后向视的标记符号的编码信息, 则 在步骤 200之前还包括对用于标记前后向视的标记符号信息进行解析的过 程, 具体为根据当前宏块周边的一个或多个宏块的标记符号建立上下文模 型, 对所述标记符号的标识信息进行解析, 所述标记符号的标识信息为将 所述当前宏块的标记符号与其周边的一个或多个宏块的标记符号进行异 或处理的结果信息; 解析出标记符号的标识信息后, 再对解析结果进行异 或处理, 得到用于标记前后向视的标记符号信息。
本实施提供的视频解码方法中, 通过解析当前宏块的对应宏块的偏移 量信息, 得到对应宏块在坐标系中的位置信息, 进而应用该对应宏块的运 动信息作为当前宏块的运动信息, 提高了解码效率。
本发明实施例还提供一种视频编码方法, 包括如下步骤: 步骤 300 , 将当前宏块的用于标记前后向视的标记符号与其周边的一 个或多个宏块的标记符号进行异或处理;
编码端通过前述或现有的判断条件, 确定选用当前宏块的前向视或后 向视的参考图像中的对应宏块, 应用选定的对应宏块的运动信息作为当前 宏块的运动信息; 而且标记符号可以标识选用的是前向视或后向视的参考 图像, 编码端将选用向视的参考图像的标记符号与其周边的一个或多个宏 块的标记符号作异或处理, 等待编码。
步骤 301 , 根据所述周边的一个或多个宏块的标记符号建立上下文模 型, 应用所述上下文模型对所述经过异或处理的标记符号信息进行编码。
利用当前宏块周边的一个或多个宏块的标记符号建立上下文模型, 选 用的周边宏块与上述步骤中选用的宏块相同, 建立上下文模型进行上下文 自适应算术编码。
若搜索在前向视和后向视的参考图像中都要进行, 则需要编码用于标记 前后向视的标记符号, 把当前编码宏块的标记符号 "currFlag" 和周边一个或 多个宏块的标记符号 "leftFlag" 作异或操作后, 建立上下文模型进行上下文 自适应算术编码。 其伪代码如下:
uiSymbol = currFlag异或 leftFlag;
uiCtx = (leftFlag==LIST_0) ? 0 : 1;
uiCtx += (aboveFlag==LIST_0) ? 0 : 1;
write Symbol( uiSymbol, MotionSkipListXFlagCCModel.get( 0, uiCtx ) ); 本领域普通技术人员可以理解: 实现上述方法实施例的全部或部分步骤可 以通过程序指令相关的硬件来完成, 前述的程序可以存储于一计算机可读 取存储介质中, 该程序在执行时, 执行包括上述方法实施例的步骤; 而前 述的存储介质包括: ROM、 RAM, 磁碟或者光盘等各种可以存储程序代 码的介质。
视频编码装置实施例 图 8为本发明视频编码装置实施例一结构示意图, 图如 8所示, 该装 置包括第一模块 11、 第二模块 12和第三模块 13 , 其中第一模块 11用于 根据预定的搜索精度的视差矢量信息, 在相邻视的参考图像中获得与当前 宏块对应的、 与所述搜索精度大小相同的图像块; 第二模块 12用于在根 据所述图像块建立的坐标系中, 获得使得所述当前宏块编码效率最优的对 应宏块的第一偏移量信息; 第三模块 13用于对所述第一偏移量信息进行 编码。
具体地, 视频编码装置中的第一模块 11按照搜索精度视差矢量信息 在参考图像中初始指定一个图像块, 该图像块的大小与搜索精度的大小相 同; 然后第二模块 12根据该图像块在参考图像中的搜索区域内建立二维 坐标系, 通过该坐标系参考图像内的所有宏块均具有位置信息; 当按照一 定的搜索顺序找到使得当前宏块的编码效率最优的对应宏块后, 得到该对 应宏块的第一偏移量信息, 即相对于坐标原点的偏移量信息; 第三模块 13 对第一偏移量信息进行编码,进一步地,第三模块 13包括第一子模块 131、 第二子模块 132和第三子模块 133 , 当第一子模块 131确定完当前宏块的 周边块例如左边块和上边块各自所在宏块在参考图像中对应的宏块的第 二偏移量信息和第三偏移量信息后, 第二子模块 132用于根据第二偏移量 信息和第三偏移量信息获得编码上下文信息; 最后, 第三子模块 133用于 应用编码上下文信息对第一偏移量信息进行编码处理。
在本视频编码装置的实施例一中, 还包括第四模块 14用于对用于标 记前后向视的标记符号信息进行编码, 具体地第四模块 14包括第八子模 块 141和第九子模块 142 , 其中第八子模块 141将所述当前宏块的用于标 记前后向视标记符号与其周边的一个或多个宏块的标记符号进行异或处 理后, 第九子模块 142根据所述周边的一个或多个宏块的标记符号建立上 下文模型, 对经过异或处理的标记符号信息进行编码。
图 9为本发明视频编码装置实施例二结构示意图, 图如 9所示, 本实 施例提供的视频编码装置与上述视频编码装置实施例一的不同之处在于, 其中第三模块 13包括第四子模块 134、 第五子模块 135、 第六子模块 136 和第七子模块 137 , 其中第四子模块 134在确定完当前宏块的周边块例如 左边块和上边块各自所在宏块在参考图像中对应的宏块的第二偏移量信 息和第三偏移量信息后, 第五子模块 135将第二偏移量信息和第三偏移量 信息的对应分量进行平均处理, 并应用平均处理结果对第一偏移量信息进 行预测, 得到预测残差信息; 在第六子模块 136根据第二偏移量信息和第 三偏移量信息获得编码上下文信息后, 第七子模块 137应用所述编码上下 文信息对该预测残差信息进行编码处理。
上述各视频编码装置实施例, 通过选择适当的搜索区域的坐标原点, 建立搜索区域内每个块的纵横坐标位置信息; 利用当前编码宏块周边块的 信息, 作为编码当前宏块在相邻视的参考图像中对应宏块的位置偏移信息 的上下文, 编码当前宏块的偏移量, 提高了编码效率。
视频解码装置实施例
图 10为本发明视频解码装置实施例一结构示意图, 如图 10所示, 该 装置包括: 第五模块 21、 第六模块 22、 第七模块 23和第八模块 24 , 其中 第五模块 21用于解析接收到的码流信息, 得到与当前宏块相邻视的参考 图像中, 与所述当前宏块对应的宏块的第一偏移量信息; 第六模块 22用 于根据视差矢量信息, 在相邻视的参考图像中获得与当前宏块对应的图像 块; 第七模块 23用于在根据所述图像块建立的参考图像搜索区域的坐标 系中, 根据所述第一偏移量信息获得与所述当前宏块对应的宏块的坐标信 息; 第八模块 24用于根据所述坐标信息获得与所述当前宏块对应的宏块 的运动信息, 应用所述运动信息进行运动补偿。
具体地, 装置中的第五模块 21接收到码流信息后, 解析得到当前解 码宏块在参考图像中的对应宏块的偏移量信息, 第七模块 23根据第六模 块 22中找到的图像块在参考图像的搜索区域范围内建立二维坐标系, 得 到对应宏块的坐标信息; 第八模块 24在已经解码的参考图像的所有宏块 的运动信息中取出对应宏块的运动信息, 作为当前宏块的运动信息进行运 动补偿。
进一步地, 第五模块 21包括第十子模块 211、 第十一子模块 212和第 十二子模块 213 , 其中第十子模块 211用于确定当前宏块的周边块例如左 边块和上边块各自所在宏块在参考图像中对应的宏块的第二偏移量信息 和第三偏移量信息; 第十一子模块 212用于根据第二偏移量信息和第三偏 移量信息获得解码上下文信息; 第十二子模块 213用于应用所述解码上下 文信息解析得到第一偏移量信息。
该装置还包括第九模块 25用于对用于标记前后向视的标记符号信息 进行解析, 在接收到码流信息后, 应用第九模块 25对码流信息中的标记 符号信息进行解析, 确定当前解码宏块的对应宏块位于哪个向视的参考图 像中。
图 11为本发明视频解码装置实施例二结构示意图, 如图 11所示, 与 视频解码装置实施例一不同之处在于第五模块 21包括第十三子模块 214、 第十四子模块 215、 第十五子模块 216和第十六子模块 217 , 其中第十三 子模块 214用于确定所述当前宏块的周边块例如左边块和上边块各自所在 宏块在所述参考图像中对应的宏块的第二偏移量信息和第三偏移量信息; 第十四子模块 215用于根据所述第二偏移量信息和第三偏移量信息获得解 码上下文信息; 第十五子模块 216用于根据所述解码上下文信息解析得到 所述对应宏块的预测残差信息; 第十六子模块 217用于将所述第二偏移量 信息和第三偏移量信息的对应分量进行平均处理, 根据处理结果和所述预 测残差信息获得所述对应宏块的第一偏移量信息。
在上述各实施提供的视频解码装置中, 通过解析当前宏块的对应宏块 的偏移量信息, 得到对应宏块在坐标系中的位置信息, 进而应用该对应宏 块的运动信息作为当前宏块的运动信息, 提高了解码效率。 视频处理***实施例
图 12为本发明视频处理***实施例一结构示意图, 如图 12所示, 该 ***包括视频编码装置 1和视频解码装置 2 , 其中视频编码装置 1包括第 一模块 11、 第二模块 12和第三模块 13 , 其中第一模块 1 1用于根据预定 的搜索精度的视差矢量信息, 在相邻视的参考图像中获得与当前宏块对应 的、 与所述搜索精度大小相同的图像块; 第二模块 12用于在根据所述图 像块建立的坐标系中, 获得使得所述当前宏块编码效率最优的对应宏块的 第一偏移量信息; 第三模块 13用于对所述第一偏移量信息进行编码。
具体地, 视频编码装置 1中的第一模块 11按照搜索精度视差矢量信 息在参考图像中初始指定一个图像块, 该图像块的大小与搜索精度的大小 相同; 然后第二模块 12根据该图像块在参考图像中的搜索区域内建立二 维坐标系, 通过该坐标系参考图像内的所有宏块均具有位置信息; 当按照 一定的搜索顺序找到使得当前宏块的编码效率最优的对应宏块后, 得到该 对应宏块的第一偏移量信息, 即相对于坐标原点的偏移量信息; 第三模块 13对第一偏移量信息进行编码, 进一步地, 第三模块 13包括第一子模块 131、 第二子模块 132和第三子模块 133 , 当第一子模块 131确定完当前宏 块的周边块例如左边块和上边块各自所在宏块在参考图像中对应的宏块 的第二偏移量信息和第三偏移量信息后, 第二子模块 132用于根据第二偏 移量信息和第三偏移量信息获得编码上下文信息; 最后, 第三子模块 133 用于应用编码上下文信息对第一偏移量信息进行编码处理。
在本视频处理***的实施例一中,视频编码装置 1还包括第四模块 14 用于对用于标记前后向视的标记符号信息进行编码, 具体地第四模块 14 包括第八子模块 141和第九子模块 142 , 其中第八子模块 141将所述当前 宏块的用于标记前后向视标记符号与其周边的一个或多个宏块的标记符 号进行异或处理后, 第九子模块 142根据所述周边的一个或多个宏块的标 记符号建立上下文模型, 对经过异或处理的标记符号信息进行编码。 视频解码装置 2包括第五模块 21、 第六模块 22、 第七模块 23和第八 模块 24, 其中第五模块 21用于解析接收到的码流信息, 得到与当前宏块 相邻视的参考图像中, 与所述当前宏块对应的宏块的第一偏移量信息; 第 六模块 22用于根据视差矢量信息, 在相邻视的参考图像中获得与当前宏 块对应的图像块; 第七模块 23用于在根据所述图像块建立的参考图像搜 索区域的坐标系中, 根据所述第一偏移量信息获得与所述当前宏块对应的 宏块的坐标信息; 第八模块 24用于根据所述坐标信息获得与所述当前宏 块对应的宏块的运动信息, 应用所述运动信息进行运动补偿。
具体地, 视频解码装置 2中的第五模块 21接收到码流信息后, 解析 得到当前解码宏块在参考图像中的对应宏块的偏移量信息, 第七模块 23 根据第六模块 22中找到的图像块在参考图像的搜索区域范围内建立二维 坐标系, 得到对应宏块的坐标信息; 第八模块 24在已经解码的参考图像 的所有宏块的运动信息中取出对应宏块的运动信息, 作为当前宏块的运动 信息进行运动补偿。
进一步地, 第五模块 21包括第十子模块 211、 第十一子模块 212和第 十二子模块 213 , 其中第十子模块 211用于确定当前宏块的周边块例如左 边块和上边块各自所在宏块在参考图像中对应的宏块的第二偏移量信息 和第三偏移量信息; 第十一子模块 212用于根据第二偏移量信息和第三偏 移量信息获得解码上下文信息; 第十二子模块 213用于应用所述解码上下 文信息解析得到第一偏移量信息。
视频解码装置 2还包括第九模块 25用于对用于标记前后向视的标记 符号信息进行解析, 在接收到码流信息后, 首先判断是否存在标记符号的 编码信息, 人若存在, 则应用第九模块 25对标记符号信息进行解析, 确 定当前解码宏块的对应宏块位于哪个向视的参考图像中。
图 13为本发明视频处理***实施例二结构示意图, 如图 13所示, 该 ***包括视频编码装置 1和视频解码装置 2 , 其中视频编码装置 1包括第 一模块 1 1、 第二模块 12和第三模块 13 , 其中第一模块 1 1用于根据预定 的搜索精度的视差矢量信息, 在相邻视的参考图像中获得与当前宏块对应 的、 与所述搜索精度大小相同的图像块; 第二模块 12用于在根据所述图 像块建立的坐标系中, 获得使得所述当前宏块编码效率最优的对应宏块的 第一偏移量信息; 第三模块 13用于对所述第一偏移量信息进行编码。
具体地, 视频编码装置 1中的第一模块 11按照搜索精度视差矢量信 息在参考图像中初始指定一个图像块, 该图像块的大小与搜索精度的大小 相同; 然后第二模块根据该图像块在参考图像中的搜索区域内建立二维坐 标系, 通过该坐标系参考图像内的所有宏块均具有位置信息; 当按照一定 的搜索顺序找到使得当前宏块的编码效率最优的对应宏块后, 得到该对应 宏块的第一偏移量信息, 即相对于坐标原点的偏移量信息; 第三模块 13 对第一偏移量信息进行编码, 进一步地第三模块 3包括第四子模块 134、 第五子模块 135、 第六子模块 136和第七子模块 137 , 其中第四子模块 134 在确定完当前宏块的周边块例如左边块和上边块各自所在宏块在参考图 像中对应的宏块的第二偏移量信息和第三偏移量信息后, 第五子模块 135 将第二偏移量信息和第三偏移量信息的对应分量进行平均处理, 并应用平 均处理结果对第一偏移量信息进行预测, 得到预测残差信息; 在第六子模 块 136根据第二偏移量信息和第三偏移量信息获得编码上下文信息后, 第 七子模块 137应用所述编码上下文信息对该预测残差信息进行编码处理。
在本视频处理***的实施例二中,视频编码装置 1还包括第四模块 14 用于对用于标记前后向视的标记符号信息进行编码, 具体地第四模块 14 包括第八子模块 141和第九子模块 142 , 其中第八子模块 141将所述当前 宏块的用于标记前后向视标记符号与其周边的一个或多个宏块的标记符 号进行异或处理后, 第九子模块 142根据所述周边的一个或多个宏块的标 记符号建立上下文模型, 对经过异或处理的标记符号信息进行编码。
视频解码装置 2包括第五模块 21、 第六模块 22、 第七模块 23和第八 模块 24, 其中第五模块 21用于解析接收到的码流信息, 得到与当前宏块 相邻视的参考图像中, 与所述当前宏块对应的宏块的第一偏移量信息; 第 六模块 22用于根据视差矢量信息, 在相邻视的参考图像中获得与当前宏 块对应的图像块; 第七模块 23用于在根据所述图像块建立的参考图像搜 索区域的坐标系中, 根据所述第一偏移量信息获得与所述当前宏块对应的 宏块的坐标信息; 第八模块 24用于根据所述坐标信息获得与所述当前宏 块对应的宏块的运动信息, 应用所述运动信息进行运动补偿。
具体地, 装置中的第五模块 21接收到码流信息后, 解析得到当前解 码宏块在参考图像中的对应宏块的偏移量信息, 第七模块 23根据第六模 块 22中找到的图像块在参考图像的搜索区域范围内建立二维坐标系, 得 到对应宏块的坐标信息; 第八模块 24在已经解码的参考图像的所有宏块 的运动信息中取出对应宏块的运动信息, 作为当前宏块的运动信息进行运 动补偿。
进一步地, 第五模块 21包括第十三子模块 214、 第十四子模块 215、 第十五子模块 216和第十六子模块 217 , 其中第十三子模块 214用于确定 所述当前宏块的周边块例如左边块和上边块各自所在宏块在所述参考图 像中对应的宏块的第二偏移量信息和第三偏移量信息; 第十四子模块 215 用于根据所述第二偏移量信息和第三偏移量信息获得解码上下文信息; 第 十五子模块 216用于根据所述解码上下文信息解析得到所述对应宏块的预 测残差信息; 第十六子模块 217用于将所述第二偏移量信息和第三偏移量 信息的对应分量进行平均处理, 根据处理结果和所述预测残差信息获得所 述对应宏块的第一偏移量信息。
该装置还包括第九模块 25用于对用于标记前后向视的标记符号信息 进行解析,在接收到码流信息后,首先判断是否存在标记符号的编码信息, 人若存在, 则应用第九模块 25对标记符号信息进行解析, 确定当前解码 宏块的对应宏块位于哪个向视的参考图像中。 上述各实施例提供的视频处理***中, 视频编码装置通过选择适当的 搜索区域的坐标原点, 建立搜索区域内每个块的纵横坐标位置信息; 利用 当前编码宏块周边块的信息, 作为编码当前宏块在相邻视的参考图像中对 应宏块的位置偏移信息的上下文, 编码当前宏块的偏移量, 提高了编码效 率; 视频解码装置通过解析当前宏块的对应宏块的偏移量信息, 得到对应 宏块在坐标系中的位置信息, 进而应用该对应宏块的运动信息作为当前宏 块的运动信息, 提高了解码效率。
最后应说明的是: 以上实施例仅用以说明本发明的技术方案, 而非对 其限制; 尽管参照前述实施例对本发明进行了详细的说明, 本领域的普通 技术人员应当理解: 其依然可以对前述各实施例所记载的技术方案进行修 改, 或者对其中部分技术特征进行等同替换; 而这些修改或者替换, 并不 使相应技术方案的本质脱离本发明各实施例技术方案的精神和范围。

Claims

权 利 要 求 书
1、 一种视频编码方法, 其特征在于包括:
根据视差矢量信息, 在相邻视的参考图像中获得与当前宏块对应的图 像块;
根据所述图像块建立该图像块所在参考图像搜索区域的坐标系; 在所述搜索区域内查找使所述当前宏块编码效率最优的对应宏块, 并 获得所述对应宏块在所述坐标系中的第一偏移量信息;
对所述第一偏移量信息进行编码。
2、 根据权利要求 1所述的视频编码方法, 其特征在于所述根据所述 图像块建立该图像块所在参考图像搜索区域的坐标系具体为:
以所述图像块或以该图像块所在宏块的首个图像块作为所述参考图 像搜索区域的坐标系的坐标原点。
3、 根据权利要求 1所述的视频编码方法, 其特征在于对所述第一偏 移量信息进行编码具体为:
确定所述当前宏块的周边块各自所在宏块在所述参考图像中对应的 宏块的偏移量信息;
根据所述周边块各自所在宏块在所述参考图像中对应的宏块的偏移 量信息获得编码上下文信息;
应用所述编码上下文信息对所述第一偏移量信息进行编码处理。
4、 根据权利要求 3所述的视频编码方法, 其特征在于应用所述编码 上下文信息对所述第一偏移量信息进行编码处理具体为:
应用截断一元码或指数哥伦布码将所述第一偏移量信息进行二值化 处理, 得到二进制位流信息;
根据所述编码上下文信息, 对所述二进制位流进行编码处理。
5、 根据权利要求 3所述的视频编码方法, 其特征在于应用所述编码 上下文信息对所述第一偏移量信息进行编码处理具体为: 应用截断一元码或指数哥伦布码将所述第一偏移量信息编码进码流。
6、 根据权利要求 1所述的视频编码方法, 其特征在于对所述第一偏 移量信息进行编码具体为:
确定所述当前宏块的周边块各自所在宏块在所述参考图像中对应的 宏块的偏移量信息;
将所述周边块各自所在宏块在所述参考图像中对应的宏块的偏移量 信息的对应分量进行平均处理, 并应用平均处理结果对所述第一偏移量信 息进行预测, 得到预测残差信息;
根据所述周边块各自所在宏块在所述参考图像中对应的宏块的偏移 量信息获得编码上下文信息;
应用所述编码上下文信息对所述预测残差信息进行编码处理。
7、 根据权利要求 6所述的视频编码方法, 其特征在于所述应用所述 编码上下文信息对所述预测残差信息进行编码处理具体为:
应用截断一元码或指数哥伦布码将所述第一偏移量信息进行二值化 处理, 得到二进制位流信息;
根据所述编码上下文信息, 将所述二进制位流进行编码处理。
8、 根据权利要求 6所述的视频编码方法, 其特征在于所述应用所述 编码上下文信息对所述预测残差信息进行编码处理具体为:
应用截断一元码或指数哥伦布码将所述第一偏移量信息各分量编码 进码流。
9、 根据权利要求 1所述的视频编码方法, 其特征在于所述对所述第 一偏移量信息进行编码之后还包括: 对用于标记前后向视的标记符号信息 进行编码。
10、 根据权利要求 9所述的视频编码方法, 其特征在于所述对用于标 记前后向视的标记符号信息进行编码具体为:
将所述当前宏块的用于标记前后向视的标记符号与其周边的一个或 多个宏块的标记符号进行异或处理;
根据所述周边的一个或多个宏块的标记符号建立上下文模型, 应用所 述上下文模型对所述经过异或处理的标记符号信息进行编码。
11、 一种视频解码方法, 其特征在于包括:
解析接收到的码流信息, 得到与当前宏块相邻视的参考图像中, 与所 述当前宏块对应的宏块的第一偏移量信息;
根据视差矢量信息, 在相邻视的参考图像中获得与当前宏块对应的图 像块;
在根据所述图像块建立的参考图像搜索区域的坐标系中, 根据所述第 一偏移量信息获得与所述当前宏块对应的宏块的坐标信息;
根据所述坐标信息获得与所述当前宏块对应的宏块的运动信息, 应用 所述运动信息进行运动补偿。
12、 根据权利要求 11所述的视频解码方法, 其特征在于所述解析接 收到的码流信息, 得到与当前宏块相邻视的参考图像中, 与所述当前宏块 对应的宏块的第一偏移量信息包括:
确定所述当前宏块的周边块各自所在宏块在所述参考图像中对应的 宏块的偏移量信息;
根据所述周边块各自所在宏块在所述参考图像中对应的宏块的偏移 量信息获得解码上下文信息;
应用所述解码上下文信息解析得到所述第一偏移量信息。
13、 根据权利要求 12所述的视频解码方法, 其特征在于所述应用所 述解码上下文信息解析得到所述第一偏移量信息具体为:
应用截断一元码或指数哥伦布码, 根据所述解码上下文信息解析得到 所述第一偏移量信息。
14、 根据权利要求 11所述的视频解码方法, 其特征在于所述解析接 收到的码流信息, 得到与当前宏块相邻视的参考图像中, 与所述当前宏块 对应的宏块的第一偏移量信息包括:
确定所述当前宏块的周边块各自所在宏块在所述参考图像中对应的 宏块的偏移量信息;
根据所述周边块各自所在宏块在所述参考图像中对应的宏块的偏移 量信息获得解码上下文信息;
根据所述解码上下文信息解析得到所述对应宏块的预测残差信息; 将所述周边块各自所在宏块在所述参考图像中对应的宏块的偏移量 信息的对应分量进行平均处理, 根据处理结果和所述预测残差信息获得所 述对应宏块的第一偏移量信息。
15、 根据权利要求 14所述的视频解码方法, 其特征在于所述根据所 述解码上下文信息解析得到所述对应宏块的预测残差信息具体为:
应用截断一元码或指数哥伦布码, 根据所述解码上下文信息解析得到 所述第一偏移量信息。
16、 根据权利要求 1 1至 15所述的任一视频解码方法, 其特征在于所 述解析接收到的码流信息, 得到与当前宏块相邻视的参考图像中, 与所述 当前宏块对应的宏块的第一偏移量信息之前还包括: 对用于标记前后向视 的标记符号信息进行解析。
17、 根据权利要求 16所述的视频解码方法, 其特征在于所述对用于 标记前后向视的标记符号信息进行解析包括:
根据所述当前宏块周边的一个或多个宏块的标记符号建立上下文模 型, 对所述标记符号的标识信息进行解析, 所述标记符号的标识信息为将 所述当前宏块的标记符号与其周边的一个或多个宏块的标记符号进行异 或处理的结果信息;
对解析结果进行异或处理, 得到用于标记前后向视的标记符号信息。
18、 一种视频编码装置, 其特征在于包括:
第一模块, 用于根据预定的搜索精度的视差矢量信息, 在相邻视的参 考图像中获得与当前宏块对应的、 与所述搜索精度大小相同的图像块; 第二模块, 用于在根据所述图像块建立的坐标系中, 获得使得所述当 前宏块编码效率最优的对应宏块的第一偏移量信息;
第三模块, 用于对所述第一偏移量信息进行编码。
19、 根据权利要求 18所述的视频编码装置, 其特征在于所述第三模 块包括:
第一子模块, 用于确定所述当前宏块的周边块各自所在宏块在所述参 考图像中对应的宏块的偏移量信息;
第二子模块, 用于根据所述周边块各自所在宏块在所述参考图像中对 应的宏块的偏移量信息获得编码上下文信息;
第三子模块, 用于应用所述编码上下文信息对所述第一偏移量信息进 行编码处理。
20、 根据权利要求 18所述的视频编码装置, 其特征在于所述第三模 块包括:
第四子模块, 用于确定所述当前宏块的周边块各自所在宏块在所述参 考图像中对应的宏块的偏移量信息;
第五子模块, 用于将所述周边块各自所在宏块在所述参考图像中对应 的宏块的偏移量信息的对应分量进行平均处理, 并应用平均处理结果对所 述第一偏移量信息进行预测, 得到预测残差信息;
第六子模块, 用于根据所述周边块各自所在宏块在所述参考图像中对 应的宏块的偏移量信息获得编码上下文信息;
第七子模块, 用于应用所述编码上下文信息对所述预测残差信息进行 编码处理。
21、 根据权利要求 18或 19或 20所述的视频编码装置, 其特征在于 还包括第四模块, 用于对用于标记前后向视的标记符号信息进行编码。
22、 根据权利要求 21所述的视频编码装置, 其特征在于所述第四模 块包括:
第八子模块, 用于将所述当前宏块的用于标记前后向视标记符号与其 周边的一个或多个宏块的标记符号进行异或处理;
第九子模块, 用于根据所述周边的一个或多个宏块的标记符号建立上 下文模型, 对所述经过异或处理的标记符号信息进行编码。
23、 一种视频解码装置, 其特征在于包括:
第五模块, 用于解析接收到的码流信息, 得到与当前宏块相邻视的参 考图像中, 与所述当前宏块对应的宏块的第一偏移量信息;
第六模块, 用于根据视差矢量信息, 在相邻视的参考图像中获得与当 前宏块对应的图像块;
第七模块, 用于在根据所述图像块建立的参考图像搜索区域的坐标系 中, 根据所述第一偏移量信息获得与所述当前宏块对应的宏块的坐标信 息;
第八模块, 用于根据所述坐标信息获得与所述当前宏块对应的宏块的 运动信息, 应用所述运动信息进行运动补偿。
24、 根据权利要求 23所述的视频解码装置, 其特征在于所述第五模 块包括:
第十子模块, 用于确定所述当前宏块的周边块各自所在宏块在所述参 考图像中对应的宏块的偏移量信息;
第十一子模块, 用于根据所述周边块各自所在宏块在所述参考图像中 对应的宏块的偏移量信息获得解码上下文信息;
第十二子模块, 用于应用所述解码上下文信息解析得到所述第一偏移 量信息。
25、 根据权利要求 23所述的视频解码装置, 其特征在于所述第五模 块包括:
第十三子模块, 用于确定所述当前宏块的周边块各自所在宏块在所述 参考图像中对应的宏块的偏移量信息;
第十四子模块, 用于根据所述周边块各自所在宏块在所述参考图像中 对应的宏块的偏移量信息获得解码上下文信息;
第十五子模块, 用于根据所述解码上下文信息解析得到所述对应宏块 的预测残差信息;
第十六子模块, 用于将所述周边块各自所在宏块在所述参考图像中对 应的宏块的偏移量信息的对应分量进行平均处理, 根据处理结果和所述预 测残差信息获得所述对应宏块的第一偏移量信息。
26、 根据权利要求 23或 24或 25所述的视频解码装置, 其特征在于 还包括第九模块, 用于对用于标记前后向视的标记符号信息进行解析。
27、 一种视频处理***, 包括视频编码装置和视频解码装置, 其特征 在于, 所述视频编码装置包括:
第一模块, 用于根据预定的搜索精度的视差矢量信息, 在相邻视的参 考图像中获得与当前宏块对应的、 与所述搜索精度大小相同的图像块; 第二模块, 用于在根据所述图像块建立的坐标系中, 获得使得所述当 前宏块编码效率最优的对应宏块的第一偏移量信息;
第三模块, 用于对所述第一偏移量信息进行编码;
所述视频解码装置包括:
第五模块, 用于解析接收到的码流信息, 得到与当前宏块相邻视的参 考图像中, 与所述当前宏块对应的宏块的第一偏移量信息;
第六模块, 用于根据视差矢量信息, 在相邻视的参考图像中获得与当 前宏块对应的图像块;
第七模块, 用于在根据所述图像块建立的参考图像搜索区域的坐标系 中, 根据所述第一偏移量信息获得与所述当前宏块对应的宏块的坐标信 息;
第八模块, 用于根据所述坐标信息获得与所述当前宏块对应的宏块的 运动信息, 应用所述运动信息进行运动补偿。
28、 根据权利要求 27所述的视频处理***, 其特征在于所述第三模 块包括:
第一子模块, 用于确定所述当前宏块的周边块各自所在宏块在所述参 考图像中对应的宏块的偏移量信息;
第二子模块, 用于根据所述周边块各自所在宏块在所述参考图像中对 应的宏块的偏移量信息获得编码上下文信息;
第三子模块, 用于应用所述编码上下文信息对所述第一偏移量信息进 行编码处理。
29、 根据权利要求 28所述的视频处理***, 其特征在于所述第五模 块包括:
第十子模块, 用于确定所述当前宏块的周边块各自所在宏块在所述参 考图像中对应的宏块的偏移量信息;
第十一子模块, 用于根据所述周边块各自所在宏块在所述参考图像中 对应的宏块的偏移量信息获得解码上下文信息;
第十二子模块, 用于应用所述解码上下文信息解析得到所述第一偏移 量信息。
30、 根据权利要求 27所述的视频处理***, 其特征在于所述第三模 块包括:
第四子模块, 用于确定所述当前宏块的周边块各自所在宏块在所述参 考图像中对应的宏块的偏移量信息;
第五子模块, 用于将所述周边块各自所在宏块在所述参考图像中对应 的宏块的偏移量信息的对应分量进行平均处理, 并应用平均处理结果对所 述第一偏移量信息进行预测, 得到预测残差信息;
第六子模块, 用于根据所述周边块各自所在宏块在所述参考图像中对 应的宏块的偏移量信息获得编码上下文信息; 第七子模块, 用于应用所述编码上下文信息对所述预测残差信息进行 编码处理。
31、 根据权利要求 30所述的视频处理***, 其特征在于所述第三模 块包括:
第十三子模块, 用于确定所述当前宏块的周边块各自所在宏块在所述 参考图像中对应的宏块的偏移量信息;
第十四子模块, 用于根据所述周边块各自所在宏块在所述参考图像中 对应的宏块的偏移量信息获得解码上下文信息;
第十五子模块, 用于根据所述解码上下文信息解析得到所述对应宏块 的预测残差信息;
第十六子模块, 用于将所述周边块各自所在宏块在所述参考图像中对 应的宏块的偏移量信息的对应分量进行平均处理, 根据处理结果和所述预 测残差信息获得所述对应宏块的第一偏移量信息。
32、 根据权利要求 27至 31所述的任一视频处理***, 其特征在于所 述视频编码装置还包括: 第四模块, 用于对用于标记前后向视的标记符号 信息进行编码。
33、 根据权利要求 32所述的视频处理***, 其特征在于所述第四模 块包括:
第八子模块, 用于将所述当前宏块的用于标记前后向视标记符号与其 周边的一个或多个宏块的标记符号进行异或处理;
第九子模块, 用于根据所述周边的一个或多个宏块的标记符号建立上 下文模型, 对所述经过异或处理的标记符号信息进行编码。
34、 根据权利要求 27至 31所述的任一视频处理***, 其特征在于所 述视频解码装置还包括: 第九模块, 用于对用于标记前后向视的标记符号 信息进行解析。
35、 一种视频编码方法, 其特征在于包括: 将当前宏块的用于标记前后向视的标记符号与其周边的一个或多个 宏块的标记符号进行异或处理;
根据所述周边的一个或多个宏块的标记符号建立上下文模型, 应用所 述上下文模型对所述经过异或处理的标记符号信息进行编码。
PCT/CN2008/073291 2008-01-04 2008-12-02 视频编码、解码方法及装置和视频处理*** WO2009086761A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US12/830,126 US20100266048A1 (en) 2008-01-04 2010-07-02 Video encoding and decoding method and device, and video processing system

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN2008100028069A CN101478672B (zh) 2008-01-04 2008-01-04 视频编码、解码方法及装置和视频处理***
CN200810002806.9 2008-01-04

Related Child Applications (1)

Application Number Title Priority Date Filing Date
US12/830,126 Continuation US20100266048A1 (en) 2008-01-04 2010-07-02 Video encoding and decoding method and device, and video processing system

Publications (1)

Publication Number Publication Date
WO2009086761A1 true WO2009086761A1 (zh) 2009-07-16

Family

ID=40839295

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2008/073291 WO2009086761A1 (zh) 2008-01-04 2008-12-02 视频编码、解码方法及装置和视频处理***

Country Status (3)

Country Link
US (1) US20100266048A1 (zh)
CN (2) CN101478672B (zh)
WO (1) WO2009086761A1 (zh)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2012044709A1 (en) * 2010-10-01 2012-04-05 General Instrument Corporation Coding and decoding utilizing picture boundary padding in flexible partitioning
WO2012044707A1 (en) * 2010-10-01 2012-04-05 General Instrument Corporation Coding and decoding utilizing picture boundary variability in flexible partitioning
US9392272B1 (en) 2014-06-02 2016-07-12 Google Inc. Video coding using adaptive source variance based partitioning
US9532059B2 (en) 2010-10-05 2016-12-27 Google Technology Holdings LLC Method and apparatus for spatial scalability for video coding
US9578324B1 (en) 2014-06-27 2017-02-21 Google Inc. Video coding using statistical-based spatially differentiated partitioning

Families Citing this family (28)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101771877B (zh) * 2010-01-20 2012-07-25 李博航 信息处理方法
CN101765012B (zh) * 2010-01-20 2012-05-23 李博航 图像信息处理方法
CA2832086C (en) * 2011-04-15 2018-10-23 Blackberry Limited Methods and devices for coding and decoding the position of the last significant coefficient
CN107517384B (zh) 2011-06-16 2020-06-30 Ge视频压缩有限责任公司 解码器、编码器、解码方法、编码方法以及存储介质
USRE47366E1 (en) 2011-06-23 2019-04-23 Sun Patent Trust Image decoding method and apparatus based on a signal type of the control parameter of the current block
EP4228264A1 (en) 2011-06-23 2023-08-16 Sun Patent Trust Image decoding device, image encoding device
KR102062283B1 (ko) 2011-06-24 2020-01-03 선 페이턴트 트러스트 화상 복호 방법, 화상 부호화 방법, 화상 복호 장치, 화상 부호화 장치 및 화상 부호화 복호 장치
WO2012176464A1 (ja) 2011-06-24 2012-12-27 パナソニック株式会社 画像復号方法、画像符号化方法、画像復号装置、画像符号化装置及び画像符号化復号装置
WO2013001764A1 (ja) 2011-06-27 2013-01-03 パナソニック株式会社 画像復号方法、画像符号化方法、画像復号装置、画像符号化装置及び画像符号化復号装置
DK2887670T3 (en) * 2011-06-28 2017-09-25 Samsung Electronics Co Ltd Method and apparatus for decoding video method and apparatus for decoding video, followed by intra-prediction.
CA2837535C (en) 2011-06-28 2018-09-04 Panasonic Corporation Image decoding method, image coding method, image decoding apparatus, image coding apparatus, and image coding and decoding apparatus
US20130083856A1 (en) * 2011-06-29 2013-04-04 Qualcomm Incorporated Contexts for coefficient level coding in video compression
WO2013001767A1 (ja) 2011-06-29 2013-01-03 パナソニック株式会社 画像復号方法、画像符号化方法、画像復号装置、画像符号化装置及び画像符号化復号装置
CA2837537C (en) 2011-06-30 2019-04-02 Panasonic Corporation Image decoding method, image coding method, image decoding apparatus, image coding apparatus, and image coding and decoding apparatus
AU2012277219A1 (en) 2011-06-30 2013-09-19 Sun Patent Trust Image decoding method, image encoding method, image decoding device, image encoding device, and image encoding/decoding device
JPWO2013008438A1 (ja) 2011-07-11 2015-02-23 パナソニック インテレクチュアル プロパティ コーポレーション オブアメリカPanasonic Intellectual Property Corporation of America 復号方法及び復号装置
UA114674C2 (uk) 2011-07-15 2017-07-10 ДЖ.І. ВІДІЕУ КЕМПРЕШН, ЛЛСі Ініціалізація контексту в ентропійному кодуванні
CN103096050B (zh) * 2011-11-04 2016-08-03 华为技术有限公司 视频图像编解码的方法及装置
LT3849198T (lt) * 2011-11-07 2023-01-25 Tagivan Ii Llc Vaizdo dekodavimo būdas ir vaizdo dekodavimo įrenginys
EP2614490B1 (en) 2011-11-11 2013-12-18 Fraunhofer-Gesellschaft zur Förderung der angewandten Forschung e.V. Concept for determining a measure for a distortion change in a synthesized view due to depth map modifications
SG10201607842VA (en) * 2012-04-13 2016-11-29 Mitsubishi Electric Corp Image encoding device, image decoding device, image encoding method and image decoding method
KR101477621B1 (ko) 2012-04-15 2015-01-02 삼성전자주식회사 변환 계수 레벨의 엔트로피 부호화 및 복호화를 위한 파라메터 업데이트 방법 및 이를 이용한 비디오 부호화 및 복호화 방법 및 장치
KR101628797B1 (ko) 2012-04-20 2016-06-09 후아웨이 테크놀러지 컴퍼니 리미티드 Hevc에서 무손실 코딩의 개선된 인트라 예측
US20130336386A1 (en) * 2012-06-18 2013-12-19 Qualcomm Incorporated Sample adaptive offset (sao) coding
US10554967B2 (en) 2014-03-21 2020-02-04 Futurewei Technologies, Inc. Illumination compensation (IC) refinement based on positional pairings among pixels
US11082720B2 (en) * 2017-11-21 2021-08-03 Nvidia Corporation Using residual video data resulting from a compression of original video data to improve a decompression of the original video data
CN112261409A (zh) * 2019-07-22 2021-01-22 中兴通讯股份有限公司 残差编码、解码方法及装置、存储介质及电子装置
CN114079771B (zh) * 2020-08-14 2023-03-28 华为技术有限公司 基于小波变换的图像编解码方法及装置

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1658673A (zh) * 2005-03-23 2005-08-24 南京大学 视频压缩编解码方法
CN1939065A (zh) * 2004-03-31 2007-03-28 皇家飞利浦电子股份有限公司 视频数据的运动估计和分段
CN101094403A (zh) * 2007-07-09 2007-12-26 西安理工大学 一种使用对角线匹配准则的运动估计方法

Family Cites Families (11)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP4610195B2 (ja) * 2001-12-17 2011-01-12 マイクロソフト コーポレーション スキップマクロブロックコード化
US6900748B2 (en) * 2003-07-17 2005-05-31 Fraunhofer-Gesellschaft Zur Foerderung Der Angewandten Forschung E.V. Method and apparatus for binarization and arithmetic coding of a data value
US7778328B2 (en) * 2003-08-07 2010-08-17 Sony Corporation Semantics-based motion estimation for multi-view video coding
DE602004017689D1 (de) * 2003-11-21 2008-12-24 Samsung Electronics Co Ltd Vorrichtung und Methode zur Erzeugung von kodierten Blockanordnungen für ein Alpha-Kanalbild sowie Alpha-Kanalkodierungs und -dekodierungsvorrichtung und -methode.
KR100624432B1 (ko) * 2004-08-05 2006-09-19 삼성전자주식회사 내용 기반 적응적 이진 산술 복호화 방법 및 장치
KR100664936B1 (ko) * 2005-04-13 2007-01-04 삼성전자주식회사 코딩 효율이 향상된 컨텍스트 기반 적응적 산술 코딩 및디코딩 방법과 이를 위한 장치, 이를 포함하는 비디오 코딩및 디코딩 방법과 이를 위한 장치
KR100718134B1 (ko) * 2005-07-21 2007-05-14 삼성전자주식회사 비트율에 적응적인 영상 데이터 이진 산술 부호화/복호화장치 및 방법
US7221296B2 (en) * 2005-08-22 2007-05-22 Streaming Networks (Pvt.) Ltd. Method and system for fast context based adaptive binary arithmetic coding
US8644386B2 (en) * 2005-09-22 2014-02-04 Samsung Electronics Co., Ltd. Method of estimating disparity vector, and method and apparatus for encoding and decoding multi-view moving picture using the disparity vector estimation method
BRPI0814854A2 (pt) * 2007-08-06 2015-01-27 Thomson Licensing Métodos e aparelho para modo de salto de vídeo animado com múltiplas imagens de referência entre exibições
US7612693B2 (en) * 2008-02-27 2009-11-03 Red Hal, Inc. Difference coding adaptive context model

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN1939065A (zh) * 2004-03-31 2007-03-28 皇家飞利浦电子股份有限公司 视频数据的运动估计和分段
CN1658673A (zh) * 2005-03-23 2005-08-24 南京大学 视频压缩编解码方法
CN101094403A (zh) * 2007-07-09 2007-12-26 西安理工大学 一种使用对角线匹配准则的运动估计方法

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2012044709A1 (en) * 2010-10-01 2012-04-05 General Instrument Corporation Coding and decoding utilizing picture boundary padding in flexible partitioning
WO2012044707A1 (en) * 2010-10-01 2012-04-05 General Instrument Corporation Coding and decoding utilizing picture boundary variability in flexible partitioning
US9532059B2 (en) 2010-10-05 2016-12-27 Google Technology Holdings LLC Method and apparatus for spatial scalability for video coding
US9392272B1 (en) 2014-06-02 2016-07-12 Google Inc. Video coding using adaptive source variance based partitioning
US9578324B1 (en) 2014-06-27 2017-02-21 Google Inc. Video coding using statistical-based spatially differentiated partitioning

Also Published As

Publication number Publication date
CN103037220B (zh) 2016-01-13
US20100266048A1 (en) 2010-10-21
CN101478672A (zh) 2009-07-08
CN101478672B (zh) 2012-12-19
CN103037220A (zh) 2013-04-10

Similar Documents

Publication Publication Date Title
WO2009086761A1 (zh) 视频编码、解码方法及装置和视频处理***
EP3402205B1 (en) Image encoding/decoding method and device
US10021367B2 (en) Method and apparatus of inter-view candidate derivation for three-dimensional video coding
CN103621093A (zh) 在三维视频编码***中的纹理图像压缩方法及装置
US10230937B2 (en) Method of deriving default disparity vector in 3D and multiview video coding
EP2207351A1 (en) The video encoding and decoding method and corresponding codec based on the motion skip mode
US20090022222A1 (en) Method and device for multi-view encoding/decoding using disparity vector
TWI551121B (zh) 視訊寫碼中之視差向量改進
CN101291434A (zh) 多视编解码方法及装置
US20140002594A1 (en) Hybrid skip mode for depth map coding and decoding
US20150304681A1 (en) Method and apparatus of inter-view motion vector prediction and disparity vector prediction in 3d video coding
KR101598855B1 (ko) 입체영상 부호화 장치 및 방법
KR102480955B1 (ko) 시점 간 움직임 병합 후보 유도 방법 및 장치
KR20150043319A (ko) 비디오 신호 처리 방법 및 장치
CN113302935A (zh) 视频编码/解码***中的端运动细化
CN113508583A (zh) 基于帧内块编译的视频或图像编译
US9900620B2 (en) Apparatus and method for coding/decoding multi-view image
US20180084276A1 (en) Method of Lookup Table Size Reduction for Depth Modelling Mode in Depth Coding
EP2717572B1 (en) Encoding/decoding method and apparatus using a skip mode
US20130100245A1 (en) Apparatus and method for encoding and decoding using virtual view synthesis prediction
CN111066322B (zh) 通过透视信息进行视频编码的帧内预测
KR101761523B1 (ko) 멀티뷰 및 3차원 비디오 코딩시의 모션 정보 예측 및 상속 방법
CN105122808A (zh) 用于三维及多视图视频编码的视差向量推导的方法及装置
CN107197288B (zh) 视频全局视差向量编码方法、解码方法和装置
KR102020024B1 (ko) 가상 시점 합성 예측을 이용한 부호화/복호화 장치 및 부호화/복호화 방법

Legal Events

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

Ref document number: 08869619

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 08869619

Country of ref document: EP

Kind code of ref document: A1