WO2011099468A1 - 動きベクトル予測符号化方法,動きベクトル予測復号方法,動画像符号化装置,動画像復号装置およびそれらのプログラム - Google Patents
動きベクトル予測符号化方法,動きベクトル予測復号方法,動画像符号化装置,動画像復号装置およびそれらのプログラム Download PDFInfo
- Publication number
- WO2011099468A1 WO2011099468A1 PCT/JP2011/052603 JP2011052603W WO2011099468A1 WO 2011099468 A1 WO2011099468 A1 WO 2011099468A1 JP 2011052603 W JP2011052603 W JP 2011052603W WO 2011099468 A1 WO2011099468 A1 WO 2011099468A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- motion vector
- primary candidate
- candidate reference
- block
- decoding
- Prior art date
Links
- 230000033001 locomotion Effects 0.000 title claims abstract description 562
- 239000013598 vector Substances 0.000 title claims abstract description 493
- 238000000034 method Methods 0.000 title claims description 81
- 238000004364 calculation method Methods 0.000 claims description 35
- 238000007796 conventional method Methods 0.000 description 15
- 238000012545 processing Methods 0.000 description 15
- 238000010586 diagram Methods 0.000 description 8
- 238000013139 quantization Methods 0.000 description 6
- 230000002123 temporal effect Effects 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- FFBHFFJDDLITSX-UHFFFAOYSA-N benzyl N-[2-hydroxy-4-(3-oxomorpholin-4-yl)phenyl]carbamate Chemical compound OC1=C(NC(=O)OCC2=CC=CC=C2)C=CC(=C1)N1CCOCC1=O FFBHFFJDDLITSX-UHFFFAOYSA-N 0.000 description 1
- 230000007423 decrease Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000007274 generation of a signal involved in cell-cell signaling Effects 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/50—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
-
- 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
- H04N19/513—Processing of motion vectors
- H04N19/521—Processing of motion vectors for estimating the reliability of the determined motion vectors or motion vector field, e.g. for smoothing the motion vector field or for correcting motion vectors
-
- 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
- H04N19/513—Processing of motion vectors
- H04N19/517—Processing of motion vectors by encoding
- H04N19/52—Processing of motion vectors by encoding by predictive encoding
-
- 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
- H04N19/533—Motion estimation using multistep search, e.g. 2D-log search or one-at-a-time search [OTS]
-
- 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
- H04N19/56—Motion estimation with initialisation of the vector search, e.g. estimating a good candidate to initiate a search
Definitions
- the present invention relates to a moving picture coding technique for predictively coding a motion vector.
- the present invention relates to a motion vector predictive encoding method, a motion vector predictive decoding method, a video encoding device, a video decoding device, and a motion vector for improving motion vector prediction efficiency and improving video encoding efficiency. It is about those programs.
- This application claims priority to Japanese Patent Application No. 2010-026130 filed in Japan on February 9, 2010, the contents of which are incorporated herein by reference.
- motion vector prediction coding is performed in order to efficiently encode a motion vector.
- FIG. 11 shows a configuration example of a motion compensation unit in a conventional video encoding apparatus.
- the motion compensation unit 100 in the conventional video encoding apparatus includes a motion search unit 101, a motion vector memory 102, a motion vector prediction unit 103, and a prediction residual calculation unit 104.
- the motion search unit 101 When the video signal of the encoding target block is input, the motion search unit 101 performs a motion search by collating with the decoded signal of the encoded reference image, calculates a motion vector, and stores it in the motion vector memory 102 To do.
- the motion vector prediction unit 103 reads a motion vector used for encoding an encoded block in the vicinity of the encoding target block from the motion vector memory 102, and calculates a prediction motion vector using these as a reference motion vector. .
- the prediction residual calculation unit 104 calculates a residual between the motion vector calculated by the motion search unit 101 and the prediction motion vector calculated by the motion vector prediction unit 103, and outputs a motion vector prediction residual. This motion vector prediction residual is encoded and output as motion vector encoding information.
- FIG. 12 shows a configuration example of a motion compensation unit in a conventional video decoding device.
- the motion compensation unit 200 in the conventional video decoding device includes a motion vector calculation unit 201, a prediction signal creation unit 202, a motion vector memory 203, and a motion vector prediction unit 204.
- the motion vector calculation unit 201 generates a motion vector by adding the motion vector prediction residual decoded from the encoded stream and the prediction motion vector predicted by the motion vector prediction unit 204, and uses the motion vector as a motion vector memory.
- the data is stored in 203 and output to the prediction signal creation unit 202.
- the prediction signal creation unit 202 reads the decoded signal from the reference image that has been decoded according to the motion vector, and outputs it as the prediction signal of the decoding target block.
- the motion vector prediction unit 204 reads a motion vector used for decoding a decoded block in the vicinity of the decoding target block from the motion vector memory 203, and calculates a prediction motion vector using them as a reference motion vector.
- FIG. 13 is a diagram for explaining an example of a conventional motion vector predictive coding method.
- a motion vector encoded motion vector
- a motion vector encoded motion vector
- the median of the reference motion vector is used as a predicted motion vector, and the error (referred to as motion vector prediction residual) between the motion vector of the encoding target block and the predicted motion vector is encoded.
- motion vector prediction residual the error between the motion vector of the encoding target block and the predicted motion vector
- a coding device selects a motion vector to be used for prediction from among the reference motion vectors, and encodes a reference motion vector identifier to be used for prediction together with a motion vector prediction residual ( Non-patent document 2).
- a technique for predicting a motion vector of a block to be encoded instead of encoding a motion vector by obtaining a motion vector prediction residual a technique for predicting a motion vector by template matching (hereinafter, conventional technique c) is used. Called).
- the prior art c is a motion vector prediction method for motion compensation without encoding the motion vectors the encoding side (see Non-Patent Document 3).
- FIG. 14 is a diagram illustrating motion vector prediction based on conventional template matching.
- a template a set of pixels that are adjacent to the encoding target block and have been encoded as shown in FIG. Is called a template
- a motion search is performed for a predetermined search range on the reference image (this process is called template matching).
- template matching a region in which the region at the same position as the template on the reference image is shifted by the motion vector
- SAD Search for similarities such as Sum of Absolute Differences.
- Motion compensation is performed using the motion vector obtained thereby. Since the same processing can be performed on the decoding side using a template that is a set of decoded pixels, there is an advantage that motion compensation can be performed without encoding a motion vector.
- the prediction efficiency of a motion vector decreases when there is no reference motion vector effective for prediction in neighboring blocks. It is also conceivable to use the reference motion vectors of a large number of blocks included not only in the vicinity of the encoding target block but also in a wider range for prediction. However, if this is done by the method of the prior art, the prediction efficiency and the coding efficiency will deteriorate.
- FIG. 15 is a diagram for explaining a problem of the prior art.
- the neighboring block of the encoding target block is the boundary of the subject Obj and when there is occlusion (when the corresponding point of the neighboring block is hidden by some subject in the reference image), the subject is If it is not a rigid body, the reference motion vector of the neighboring block may not be suitable for the motion vector prediction of the encoding target block, or intra coding may be performed and the reference motion vector itself may not exist. In such a case, the prediction efficiency is deteriorated in both the conventional technique a and the conventional technique b.
- a motion vector of a block not included in a candidate may be more effective for prediction.
- the number of candidate blocks is increased, not just the nearest block.
- the reference motion vector that is not appropriate in the prior art “a” may be included in the candidates, and the prediction efficiency may deteriorate.
- the code amount of the identifier of the reference motion vector used for prediction is increased, so that the coding efficiency may be deteriorated.
- the conventional technique c is a motion vector prediction method for performing motion compensation without encoding a motion vector on the encoding side. Therefore, consider applying this to the problems of the prior art described above. That is, it is considered that a motion vector prediction residual is generated from a motion vector of a target block to be encoded obtained by generating a motion vector predictor using template matching of the conventional technique c and a normal motion search and encoded. . In this case, there are the following problems.
- the search can be performed without using the encoded motion vectors of the neighboring blocks of the encoding target block. For this reason, even when the encoded motion vector is not effective for prediction, there is a possibility that an effective predicted motion vector can be created. However, since the predicted motion vector is determined only from the template, a motion vector indicating an area unrelated to the encoding target block is used as the predicted motion vector, and the prediction efficiency may deteriorate.
- An object of the present invention is to solve the above-described problems, improve the prediction efficiency of motion vectors, and improve the encoding efficiency of moving images.
- the prediction efficiency of a motion vector means the degree of similarity between a motion vector to be predicted and a predicted motion vector. Specifically, it is assumed that the prediction efficiency is high when the length of the difference vector between these two vectors is small.
- the outline of the present invention is as follows.
- the present invention performs motion vector prediction for each block on the encoding side and decoding side by the following method.
- N a large number of encoded primary candidate blocks determined by the position of the encoding target block
- the primary candidate reference motion vectors are narrowed down to M ( ⁇ N) secondary candidate reference motion vectors according to the reliability.
- a predicted motion vector is created using M secondary candidate reference motion vectors.
- processing 1 to processing 3 are performed as preprocessing of motion vector predictive coding (processing 4 below) similar to the conventional one.
- N blocks (N is an integer greater than or equal to 2) are set as primary candidate blocks, and N primary candidate reference motion vectors are determined from the motion vectors used for encoding the primary candidate blocks.
- the top M reference candidate motion vectors (M is an integer greater than or equal to 1 and less than N) with high reliability are used as secondary candidate reference motion vectors. elect.
- a prediction motion vector of the encoding target block is calculated using the secondary candidate reference motion vector, and a residual between the motion vector obtained by the motion search of the encoding target block and the prediction motion vector is calculated as a motion vector. It encodes as encoding information.
- a median value of M secondary candidate reference motion vectors is selected, or M secondary candidate reference motion vectors are selected. Conventional methods such as selecting a secondary candidate reference motion vector having the smallest prediction residual among the vectors and encoding the motion vector identifier together with the prediction residual can be used.
- motion vectors of a large number of primary candidate blocks in a predetermined range in at least one of an encoding target image and an encoded image as well as neighboring blocks of the encoding target block To determine a primary candidate reference motion vector. Then, reliability is calculated for each primary candidate reference motion vector using encoded information or decoded information. The primary candidate reference motion vector is narrowed down according to the reliability, and the narrowed result is set as the secondary candidate reference motion vector. In the subsequent processing, using the secondary candidate reference motion vector as an input, for example, using the same method as the conventional motion vector predictive coding, a predicted motion vector is obtained, and the prediction residual between the predicted motion vector and the motion vector is coded .
- motion vectors of a large number of predetermined primary candidate blocks are set as primary candidate reference motion vectors.
- the reliability is calculated using the decoded information.
- the primary candidate reference motion vector is narrowed down according to the reliability, and the narrowed result is set as the secondary candidate reference motion vector.
- Subsequent processing uses the secondary candidate reference motion vector as input, obtains a predicted motion vector using the same method as conventional motion vector predictive decoding, and calculates the motion vector by adding the predicted motion vector to the decoded prediction residual. To do.
- a method of calculating the reliability for example, a set of encoded pixels adjacent to the encoding target block is used as a template, and an area that is spatially identical to the template on the reference image is shifted by the primary candidate reference motion vector. It is possible to use a method of calculating the similarity between the matching target region and the template and setting the similarity as the reliability.
- the reliability may be calculated using the smallness of the decoded prediction residual signal in the motion compensation of the primary candidate block from which each primary candidate reference motion vector is calculated as an index of reliability.
- the reference motion vectors are narrowed down by performing the above processes 1 to 3. This narrowing down can be realized on the decoding side without additional information from the encoding side, and the secondary candidate reference motion vector includes a motion vector effective for prediction. For this reason, prediction efficiency improves from the prior art a, b, c mentioned above.
- the entropy of the motion vector prediction residual is reduced and the code amount of the motion vector is reduced. Since the encoded data of the moving image includes the code amount of the motion vector, the encoding efficiency of the moving image is improved as compared with the method using the conventional techniques a, b, and c.
- primary candidate reference motion vectors can be set from a large number of primary candidate blocks, and these primary candidate reference motion vectors are motion vectors used for encoding. It corresponds to the motion of the subject in the moving image, and there is a high possibility that a motion vector effective for motion vector prediction is included.
- the reliability is obtained only for these primary candidate reference motion vectors, it is possible to realize high prediction efficiency with a smaller amount of calculation than in the conventional technique c.
- FIG. 1 It is a figure which shows the structural example of the moving image encoder which concerns on embodiment of this invention. It is a figure which shows the detailed structural example of the motion compensation part shown in FIG. It is a figure which shows the structural example of the moving image decoding apparatus which concerns on embodiment of this invention. It is a figure which shows the detailed structural example of the motion compensation part shown in FIG. It is a flowchart of a motion vector prediction process. It is a figure explaining the example of a setting of a primary candidate block. It is a figure explaining the example of a setting of a primary candidate block. It is a figure explaining the example of a setting of a primary candidate reference motion vector. It is a flowchart which shows an example of a reliability calculation process.
- FIG. 1 is a diagram illustrating a configuration example of a moving image encoding apparatus according to an embodiment of the present invention.
- the part of the motion compensation unit 18 is particularly different from the prior art, and the other part is H.264. It is the same as that of a conventional general video encoding apparatus used as an encoder in H.264 and others.
- the moving image encoding apparatus 1 receives a video signal to be encoded, divides a frame of the input video signal into blocks, encodes each block, and outputs the encoded data as a bit stream.
- the prediction residual signal calculation unit 10 obtains a difference between the input video signal and the prediction signal that is the output of the motion compensation unit 18, and outputs it as a prediction residual signal.
- the orthogonal transform unit 11 performs orthogonal transform such as discrete cosine transform (DCT) on the prediction residual signal and outputs a transform coefficient.
- the quantization unit 12 quantizes the transform coefficient and outputs the quantized transform coefficient.
- the information source encoding unit 13 entropy encodes the quantized transform coefficient and outputs it as a bit stream.
- the quantized transform coefficient is also input to the inverse quantization unit 14 where it is inversely quantized.
- the inverse orthogonal transform unit 15 performs inverse orthogonal transform on the transform coefficient output from the inverse quantization unit 14 and outputs a prediction residual decoded signal.
- the decoded signal calculation unit 16 adds the prediction residual decoded signal and the prediction signal output from the motion compensation unit 18 to generate a coded decoded signal of the encoding target block.
- the decoded signal is stored in the frame memory 17 for use as a motion compensation reference image in the motion compensation unit 18.
- the motion compensation unit 18 performs a motion search on the video signal of the encoding target block with reference to the reference image stored in the frame memory 17 and outputs a prediction signal of the encoding target block.
- the motion compensation unit 18 predicts a motion vector using the encoded information in order to predictively encode a motion vector as a result of motion search, a motion vector as a result of motion search, and a predicted motion. The difference from the vector is calculated, and the result is output to the information source encoding unit 13 as a motion vector prediction residual.
- the motion compensation unit 18 does not simply use the motion vector of the encoded block in the vicinity of the encoding target block but predicts the motion vector in advance with respect to the position of the encoding target block.
- a large number of encoded primary candidate blocks at relative positions are set, a primary candidate reference motion vector is set from the motion vectors used for encoding the primary candidate block, and the primary candidate reference is set.
- the motion vector reliability is calculated from the encoded information, narrowed down to a small number of secondary candidate reference motion vectors according to the reliability, and then the predicted motion vector is calculated using the secondary candidate reference motion vectors.
- the process of calculating the predicted motion vector using the secondary candidate reference motion vector can be performed using a motion vector prediction method similar to the conventional technique.
- FIG. 2 is a diagram showing a detailed configuration example of the motion compensation unit 18 shown in FIG.
- the motion compensation unit 18 includes a motion search unit 181, a motion vector memory 182, a primary candidate block motion vector reading unit 183, a reliability calculation unit 184, a reference motion vector determination unit 185, and a motion vector prediction. 186, a motion vector prediction residual calculation unit 187.
- the motion search unit 181 performs a motion search for matching a block to be coded of an input video signal with a decoded signal of a reference image that has already been coded, and a prediction signal Is generated and output, and a motion vector indicating the matching position is output.
- This motion vector is stored in the motion vector memory 182 and output to the motion vector prediction residual calculation unit 187.
- the primary candidate block motion vector reading unit 183 encodes N (N is an integer of 2 or more) primary candidate blocks at predetermined positions, which are encoded in the past and stored in the motion vector memory 182. Are set as primary candidate reference motion vectors and notified to the reliability calculation unit 184.
- the reliability calculation unit 184 uses the encoded image information (decoded signal) to quantitatively evaluate the effectiveness in motion vector prediction in the encoding target block. Calculate the confidence level to represent.
- the reference motion vector determination unit 185 selects M primary candidate reference motion vectors (M is an integer greater than or equal to 1 and less than N) from the top in descending order of reliability calculated by the reliability calculation unit 184, as secondary candidate reference motions. Elected as a vector.
- the motion vector prediction unit 186 calculates a prediction motion vector of the encoding target block using the secondary candidate reference motion vector selected by the reference motion vector determination unit 185.
- the motion vector predictor 186 calculates the motion vector predictor in the same manner as in the prior art. For example, the median value in the secondary candidate reference motion vector is used as the motion vector predictor.
- the secondary candidate reference motion vector having the closest value to the motion vector obtained by the motion search unit 181 is set as a predicted motion vector, and an identifier indicating the motion vector is added to the encoding target, and then the decoding side. It is also possible to notify.
- the motion vector prediction residual calculation unit 187 calculates a residual between the motion vector calculated by the motion search unit 181 and the prediction motion vector calculated by the motion vector prediction unit 186, and uses the calculated residual as a motion vector prediction residual. Output as difference.
- FIG. 3 is a diagram showing a configuration example of the moving picture decoding apparatus according to the embodiment of the present invention.
- the part of the motion compensation unit 25 is different from the prior art, and the other part is H.264. It is the same as that of a conventional general video decoding device used as a decoder in H.264 and others.
- the moving picture decoding apparatus 2 outputs the decoded signal of the decoded picture by inputting and decoding the bit stream encoded by the moving picture encoding apparatus 1 shown in FIG.
- the information source decoding unit 20 entropy-decodes the quantized transform coefficient of the block to be decoded based on the input bitstream and also decodes the motion vector prediction residual.
- the inverse quantization unit 21 receives a quantized transform coefficient, inversely quantizes it, and outputs a decoded transform coefficient.
- the inverse orthogonal transform unit 22 performs inverse orthogonal transform on the decoded transform coefficient and outputs a decoded prediction residual signal.
- the decoding signal calculating section 23, by adding the a prediction signal generated by the motion compensation unit 25 decodes the prediction residual signal to generate a decoded signal of the decoding target block.
- the decoded signal is output to an external device such as a display device, for use as a reference picture of motion compensation in the motion compensation unit 25, it is stored in the frame memory 24.
- the motion compensation unit 25 performs motion vector prediction using the decoded information stored in the frame memory 24, and adds the predicted motion vector and the motion vector prediction residual decoded by the information source decoding unit 20.
- a motion vector is calculated, and a reference signal in the frame memory 24 is referred to based on the motion vector to generate a prediction signal for the decoding target block.
- the motion compensation unit 25 does not simply use the motion vector of the decoded block in the vicinity of the block to be decoded, but predicts the motion vector by using a predetermined relative position with respect to the position of the block to be decoded. A large number of decoded primary candidate blocks at different positions are set, and a primary candidate reference motion vector is set from the motion vectors used for decoding the primary candidate block.
- the motion compensation unit 25 calculates the reliability of the primary candidate reference motion vector from the decoded information, narrows down to a small number of secondary candidate reference motion vectors according to the reliability, and then performs secondary candidate reference motion. A predicted motion vector is calculated using the vector. The process of calculating the predicted motion vector using the secondary candidate reference motion vector can be performed using a motion vector prediction method similar to the conventional technique.
- FIG. 4 is a diagram showing a detailed configuration example of the motion compensation unit 25 shown in FIG.
- the motion compensation unit 25 includes a motion vector calculation unit 251, a prediction signal creation unit 252, a motion vector memory 253, a primary candidate block motion vector reading unit 254, a reliability calculation unit 255, and a reference motion vector.
- a determination unit 256 and a motion vector prediction unit 257 are provided.
- the motion vector calculation unit 251 uses the motion vector prediction residual obtained by decoding the encoded bitstream and the information that the motion vector prediction unit 257 has decoded.
- the predicted motion vector predicted is added and a motion vector used for decoding is output.
- This motion vector is stored in the motion vector memory 253 and is output to the prediction signal creation unit 252.
- the prediction signal creation unit 252 reads the decoded signal at the reference image position indicated by the input motion vector and outputs it as a prediction signal of the decoding target block.
- the primary candidate block motion vector reading unit 254 decodes N (N is an integer of 2 or more) primary candidate blocks at predetermined positions, which have been decoded in the past and stored in the motion vector memory 253.
- the motion vectors are read out, these motion vectors are set as primary candidate reference motion vectors, and the reliability calculation unit 255 is notified.
- the reliability calculation unit 255 uses the decoded image information (decoded signal) for each of the N primary candidate reference motion vectors to quantitatively represent the effectiveness in motion vector prediction in the decoding target block. Calculate the degree.
- the reference motion vector determination unit 256 outputs M primary candidate reference motion vectors (M is an integer of 1 or more and less than N) from the top in the descending order of reliability calculated by the reliability calculation unit 255, as secondary candidate reference motions. Elected as a vector.
- the motion vector prediction unit 257 calculates the prediction motion vector of the decoding target block using the secondary candidate reference motion vector selected by the reference motion vector determination unit 256.
- the motion vector prediction unit 257 may calculate the motion vector predictor in the same manner as in the prior art. For example, the median value in the secondary candidate reference motion vector is used as the motion vector predictor.
- the motion vector indicated by the identifier is set as a predicted motion vector.
- motion vector prediction processing related to the present invention will be described with reference to FIGS. 5 to 10B. To do. In the following, the description will focus on the motion vector prediction process on the encoding side, but the motion vector prediction process on the decoding side is exactly the same.
- FIG. 5 shows a flowchart of motion vector prediction processing.
- the motion vector reading unit 183 (or 254) of the primary candidate block reads the motion vectors of the N primary candidate blocks from the motion vector memory 182 (or 253).
- the primary candidate block can be selected and set from an encoding (decoding) target picture, or can be selected and set from an encoded picture.
- FIG. 6A shows an example in which an encoded block in the encoding target picture 3 is set as a primary candidate block.
- encoded blocks B1 ⁇ B10 numerous in the neighborhood of the encoding target block 31 (10 pieces in this example) is set as the primary candidate blocks.
- the decoded block in the same position as the primary candidate blocks used in encoding side can be set as the primary candidate blocks.
- FIG. 6B shows an example in which the encoded block in the encoded picture 30 is set as the primary candidate block. For example, if the position of the block B11 (corresponding position block) in the encoded picture 30 is the same as the position of the encoding target block 31 in the encoding target picture 3, the block B11 and some of its surroundings Blocks B12 to B19 are set as primary candidate blocks.
- encoded picture block may be set as a primary candidate block, but a plurality of encoded picture blocks may be set as primary candidate blocks.
- the primary candidate block motion vector reading unit 183 determines whether each of the N primary candidate blocks is a block of an encoded picture (or a decoded picture). If the primary candidate block is an encoded picture block, the motion vector reading unit 183 of the primary candidate block executes the next step S3.
- Step S3 If the primary candidate block is an encoded picture, the motion vector itself used for encoding is not used as the primary candidate reference motion vector, but corresponds to the reference image of the encoding target picture. Process motion vectors.
- the motion vector is processed as follows.
- V V c ⁇ (T r ⁇ T e ) / (T r2 ⁇ T c )
- the “time” of the above picture is the display time information of the picture, Any device may be used as long as it indicates the relative time relationship of pictures, such as POC (Picture Order Count) defined in the H.264 standard.
- POC Picture Order Count
- FIG. 7 shows a setting example of the primary candidate reference motion vector for the primary candidate block of the encoded picture.
- the encoded picture may be any picture as long as it is an inter picture (a picture that is encoded by performing motion compensation). For example, H.M. In the case of a B picture in H.264, there are cases where the preceding and following P pictures can be used.
- P2 is an encoding target picture 3 and P4 is an encoded picture 30 in which a primary candidate block exists. Further, it is assumed that P1 and P4 are P pictures and P2 and P3 are B pictures among P1 to P4.
- the encoding target picture 3 of P2 uses the encoded picture 30 of P4 as a reference image, and the encoded picture 30 of P4 uses P1 as a reference image.
- the motion vector V p4 (dotted arrow in the figure) of the primary candidate block in the encoded picture 30 of P4 is for the encoded picture of P1. Therefore, from the temporal distance L1 between P1 and P4 and the temporal distance L2 between P2 and P4 and the reference directionality, the primary candidate reference motion vector V (solid arrow in the figure) is calculated by the following equation.
- the motion vector reading unit 183 (or 254) of the primary candidate block sets the motion vector obtained from the N primary candidate blocks by the above processing as the primary candidate reference motion vector, and the reliability calculation unit 184. (Or 255).
- the reliability calculation unit 184 calculates the reliability for each of the set N primary candidate reference motion vectors using the encoded information.
- the reliability is a quantitative expression of the effectiveness of the primary candidate reference motion vector in the motion vector prediction in the encoding (decoding) target block. This reliability is calculated for only the N primary candidate reference motion vectors using only information that has already been decoded at the time when decoding starts on the decoding side.
- FIG. 8 is a flowchart showing an example of the reliability calculation process.
- FIG. 9 is a diagram for explaining how to obtain reliability using template matching.
- the template 32 is a set of encoded pixels adjacent to the encoding target block 31 (in this example, an inverted L-shaped region composed of a pixel group on the left and the upper side of the encoding target block 31).
- the width (thickness) of the inverted L-shaped region is, for example, about 2 pixels, but may be 1 pixel or 3 pixels or more.
- the reference image 4 is an encoded or decoded picture.
- the corresponding position block 41 in the reference image 4 is a block at the same position as the position of the encoding target block 31 in the encoding target picture 3.
- step S51 an attempt is made to calculate the reliability of an area spatially the same as the template 32 (an inverted L-shaped area adjacent to the corresponding position block 41) on the reference image 4.
- a region shifted by the existing primary candidate reference motion vector V i is obtained, and this is acquired as the matching target region 42.
- step S52 the similarity between the template 32 of the encoding target block 31 and the matching target area 42 in the reference image 4 is calculated and set as the reliability of the primary candidate reference motion vector V i .
- SAD Sud of Absolute Differences
- the reliability index used in the reliability calculation unit 184 may be any other index as long as it indicates the similarity between the template 32 and the matching target area 42.
- SSD Sum of Squared Differences
- SATD Sud of Absolute Transformed Differences
- the template 32 Since the template 32 has a high correlation with the image signal of the encoding target block 31, it is possible to specify a secondary candidate reference block effective for motion vector prediction using the similarity.
- a method using the small size of the decoded prediction residual signal in the motion compensation of the primary candidate block can be used.
- the primary candidate block is highly likely to be a subject boundary, and therefore is highly likely to be a motion vector that is not effective for motion vector prediction.
- the smaller the decoded prediction residual signal the higher the reliability.
- an absolute value sum, a square sum, etc. of the decoded prediction residual signal can be used as an indicator of the smallness of the decoded prediction residual signal.
- Step S6 In the reference motion vector determination unit 185 (or 256), M primary candidate reference motion vectors (1 ⁇ M ⁇ 1) are determined based on each primary candidate reference motion vector • BR> pump reliability information. N) The secondary candidate reference motion vector is narrowed down.
- FIG. 10A is a flowchart of the reference motion vector determination process.
- the reference motion vector determination unit 185 arranges the reliability of the primary candidate reference motion vectors calculated by the reliability calculation unit 184 in descending order, and outputs the primary candidate reference motion vectors having the highest M reliability as the secondary. Set as a candidate reference motion vector.
- FIG. 10B is a flowchart of another reference motion vector determination process, and shows an example of a reference motion vector determination process when a case where the number of primary candidate reference motion vectors is less than M is considered.
- the secondary candidate reference motion vector is determined as a reference motion vector as follows.
- step S62 it is determined whether or not the number N of primary candidate reference motion vectors is larger than M. If N is greater than M, the process proceeds to step S63, and the upper M primary candidate reference motion vectors are set as secondary candidate reference motion vectors for reliability as in step S61 described above. If the number N of primary candidate reference motion vectors that can actually be used is not greater than M, the process proceeds to step S64, and N primary candidate reference motion vectors are set as secondary candidate reference motion vectors. .
- the motion vector prediction unit 186 uses the secondary candidate reference motion vector selected by the reference motion vector determination unit 185 to create a prediction motion vector of the encoding target block.
- the important point in the present invention is that a motion vector prediction residual for calculating a motion vector prediction residual using a highly reliable secondary candidate reference motion vector by narrowing down a large number of primary candidate reference motion vectors by reliability. The point is to find a vector. Therefore, the process for obtaining the predicted motion vector from the secondary candidate reference motion vector may be the same as the process of the motion vector predicting unit 103 (or 204) of the prior art described with reference to FIGS. However, the processing does not necessarily have to be the same as that in the prior art, and the present invention can also be implemented by obtaining a predicted motion vector by different processing.
- the primary candidate reference motion vector can also be set as follows.
- the motion vectors of the encoded block are candidates, the motion vectors within a predetermined range for these motion vectors are also used as primary candidate reference motion vectors.
- the predetermined range is a range of ⁇ 1 in the X and Y directions
- the motion vector of (10, 20) (9, 20), (11, 20), (10, 19), (10, 21), (9, 19), (9, 21), (11, 19), (11, 21) motion vectors are also candidates.
- a total of nine primary candidate reference motion vectors are candidates for the motion vector of one encoded block. If the number of motion vectors of an encoded block that is initially included in the candidate is K, and all the K surroundings are also included in the candidate, 9 ⁇ K primary candidate reference motion vectors are used. However, if it is common with the decoding side, the periphery of the motion vectors of all the encoded blocks is not included in the candidates but may be a part.
- the effect of such setting is that the motion vector prediction efficiency is further improved by taking into account the periphery of the motion vector of the encoded block.
- the above-described motion vector predictive encoding process and motion vector predictive decoding process can also be realized by a computer and a software program.
- the program can be recorded on a computer-readable recording medium or provided through a network.
- the present invention can be used for, for example, moving picture coding for predictive coding of motion vectors. According to the present invention, motion vector prediction efficiency is improved, and moving picture encoding efficiency is improved.
- Video coding apparatus 2 Video decoding apparatus 10 Prediction residual signal calculation part 11 Orthogonal transformation part 12 Quantization part 13 Information source coding part 14, 21 Inverse quantization part 15, 22 Inverse orthogonal transformation part 16 Decoding signal calculation Unit 17, 24 frame memory 18, 25 motion compensation unit 181 motion search unit 182, 253 motion vector memory 183, 254 primary candidate block motion vector reading unit 184, 255 reliability calculation unit 185, 256 reference motion vector determination unit 186 , 257 Motion vector prediction unit 187 Motion vector prediction residual calculation unit 20 Information source decoding unit 23 Decoded signal calculation unit 251 Motion vector calculation unit 252 Prediction signal generation unit
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
Description
本願は,2010年2月9日に日本へ出願された日本特願2010-026130号に対して優先権を主張し,その内容をここに援用する。
(a)メディアン予測符号化(H.264など)〔以下,従来技術aという〕
(b)参照動きベクトル指定による予測符号化〔以下,従来技術bという〕
図13は,従来の動きベクトルの予測符号化方式の例を説明する図である。従来技術aおよび従来技術bでは,動きベクトルを符号化(復号も同様)する際に,図13に示すような符号化対象ブロックの近隣の符号化済みブロックの動きベクトル(符号化済み動きベクトル)を参照動きベクトルとして用いて予測を行って,動きベクトルの符号化を行う。
(1)符号化対象画像および符号化済み画像の少なくとも一方において,符号化対象ブロックの位置によって定まる,符号化済みの多数(N個)の1次候補ブロックの動きベクトルを,1次候補参照動きベクトルとして利用する。
(2)復号側で符号化(復号)対象ブロックを復号開始する時点ですでに復号済みの情報のみを利用して,各1次候補参照動きベクトルがどれだけ予測に適しているかを示す評価値(以下,信頼度)を求める。
(3)信頼度に応じて,1次候補参照動きベクトルをM(<N)個の2次候補参照動きベクトルに絞り込む。
(4)M個の2次候補参照動きベクトルを利用して予測動きベクトルを作成する。
動きベクトル予測残差算出部187は,動き探索部181が算出した動きベクトルと,動きベクトル予測部186が算出した予測動きベクトルとの残差を算出し,算出された残差を動きベクトル予測残差として出力する。
最初に,1次候補ブロックの動きベクトル読み出し部183(または254)は,N個の1次候補ブロックの動きベクトルを,動きベクトルメモリ182(または253)から読み出す。
1次候補ブロックの動きベクトル読み出し部183は,N個の1次候補ブロックの各々について,符号化済みピクチャ(または復号済みピクチャ)のブロックであるかどうかを判定する。1次候補ブロックが符号化済みピクチャのブロックである場合には,1次候補ブロックの動きベクトル読み出し部183は,次のステップS3を実行する。
1次候補ブロックが符号化済みピクチャのものであれば,その符号化に利用された動きベクトルそのものを1次候補参照動きベクトルとするのではなく,符号化対象ピクチャの参照画像に対応するように動きベクトルを加工する。
上記のピクチャの「時刻」は,ピクチャの表示時刻情報や,H.264の規格で定義されているPOC(Picture Order Count )など,ピクチャの相対的な時間関係を示すものであればどのようなものでもよい。
すなわち,前述した1次候補参照動きベクトルVを算出する式「V=Vc ×(Tr -Te )/(Tr2-Tc )」に当てはめると,Vp4=Vc ,ピクチャP2の時刻がTe ,ピクチャP4の時刻がTc =Tr ,ピクチャP1の時刻がTr2となる。この式により,1次候補ブロックの動きベクトルVp4を加工して,1次候補参照動きベクトルVとして用いることができるようにする。
1次候補ブロックの動きベクトル読み出し部183(または254)は,以上の処理によってN個の1次候補ブロックから得られた動きベクトルを,1次候補参照動きベクトルとして設定し,信頼度計算部184(または255)に通知する。
信頼度計算部184(または255)は,設定されたN個の1次候補参照動きベクトルの各々について,符号化済みの情報を用いて信頼度を算出する。ここで,信頼度は,符号化(復号)対象ブロックでの動きベクトル予測における1次候補参照動きベクトルの有効性を定量的に表現したものである。この信頼度は,N個の1次候補参照動きベクトルについて,復号側で符号化対象ブロックの復号を開始する時点ですでに復号済みの情報のみを利用して計算する。
次に,参照動きベクトル決定部185(または256)において,各1次候補参照動きベクト・BR>汲フ信頼度情報に基づき,N個の1次候補参照動きベクトルをM個(1≦M<N)の2次候補参照動きベクトルに絞り込む。
動きベクトル予測部186(または257)は,参照動きベクトル決定部185が選出した2次候補参照動きベクトルを利用して,符号化対象ブロックの予測動きベクトルを作成する。本発明において重要なポイントは,多数の1次候補参照動きベクトルを信頼度によって絞り込むことにより,信頼度の高い2次候補参照動きベクトルを用いて,動きベクトル予測残差を算出するための予測動きベクトルを求める点にある。したがって,2次候補参照動きベクトルから予測動きベクトルを求める処理は,図11や図12で説明した従来技術の動きベクトル予測部103(または204)の処理と同様でよい。しかし,必ずしも従来技術と同じ処理でなければならないわけではなく,異なる処理によって予測動きベクトルを求めて,本発明を実施することもできる。
2 動画像復号装置
10 予測残差信号算出部
11 直交変換部
12 量子化部
13 情報源符号化部
14,21 逆量子化部
15,22 逆直交変換部
16 復号信号算出部
17,24 フレームメモリ
18,25 動き補償部
181 動き探索部
182,253 動きベクトルメモリ
183,254 1次候補ブロックの動きベクトル読み出し部
184,255 信頼度計算部
185,256 参照動きベクトル決定部
186,257 動きベクトル予測部
187 動きベクトル予測残差算出部
20 情報源復号部
23 復号信号算出部
251 動きベクトル算出部
252 予測信号作成部
Claims (14)
- 符号化対象画像をブロックに分割し,ブロックごとに動き補償を用いて符号化する動画像符号化方式における動きベクトル予測符号化方法において,
符号化済みの参照画像を用いて,前記符号化対象画像における符号化対象ブロックの動き探索を行い,動きベクトルを算出するステップと,
前記符号化対象画像における前記符号化対象ブロックの位置に対して,予め定められた相対的な位置にあり,前記符号化対象画像中の符号化済みブロックおよび符号化済み画像における符号化済みブロックの少なくとも一方からなる複数個のブロックを1次候補ブロックとして設定し,前記1次候補ブロックの符号化に用いた動きベクトルからN個(Nは2以上の整数)の1次候補参照動きベクトルを決定するステップと,
前記1次候補参照動きベクトルの,前記符号化対象ブロックでの動きベクトル予測における有効性を定量的に表す信頼度を,前記1次候補参照動きベクトルのそれぞれについて符号化済みの画像情報を用いて算出するステップと,
前記N個の1次候補参照動きベクトルの中から,信頼度が大きい上位M個(Mは1以上かつN未満の所定の整数)の1次候補参照動きベクトルを,2次候補参照動きベクトルとして選出するステップと,
前記2次候補参照動きベクトルを用いて前記符号化対象ブロックの予測動きベクトルを算出し,前記符号化対象ブロックの動き探索で求めた前記動きベクトルと,前記予測動きベクトルとの残差を,前記動きベクトルの符号化情報として符号化するステップとを有する
動きベクトル予測符号化方法。 - 請求項1記載の動きベクトル予測符号化方法において,
N個設定すべき前記1次候補参照動きベクトルが前記M個より小さいM′個しか設定できなかった場合に,M′個の1次候補参照動きベクトルを,前記2次候補参照動きベクトルとして選出する
動きベクトル予測符号化方法。 - 請求項1または請求項2記載の動きベクトル予測符号化方法において,
前記1次候補参照動きベクトルの信頼度を算出するステップでは,前記符号化対象ブロックに隣接する符号化済み画素の集合をテンプレートとして用い,前記参照画像上で前記1次候補参照動きベクトル分だけ前記テンプレートの領域をずらした領域をマッチング対象領域として設定し,前記テンプレートの前記符号化済み画素の集合と前記マッチング対象領域における画素の集合の類似度を前記信頼度として算出する
動きベクトル予測符号化方法。 - 請求項1または請求項2記載の動きベクトル予測符号化方法において,
前記1次候補参照動きベクトルの信頼度を算出するステップでは,前記1次候補参照動きベクトルに係る1次候補ブロックの動き補償における復号予測残差信号の小ささを信頼度の指標として,前記信頼度を算出する
動きベクトル予測符号化方法。 - 請求項1から請求項4までのいずれか1項に記載の動きベクトル予測符号化方法において,
前記1次候補参照動きベクトルを決定するステップでは,前記1次候補ブロックの符号化に用いた前記動きベクトルと,当該動きベクトルの各々を基準とした所定範囲内の動きベクトルとから,前記1次候補参照動きベクトルを決定する
動きベクトル予測符号化方法。 - ブロックに分割されて符号化された動画像の復号対象画像を,ブロックごとに動き補償を用いて復号する動画像復号方式における動きベクトル予測復号方法において,
復号対象ブロックの動きベクトル予測残差を復号するステップと,
前記復号対象画像における前記復号対象ブロックの位置に対して,予め定められた相対的な位置にあり,前記復号対象画像中の復号済みブロックおよび復号済み画像における復号済みブロックの少なくとも一方からなる複数個のブロックを1次候補ブロックとして設定し,前記1次候補ブロックの復号に用いた動きベクトルからN個(Nは2以上の整数)の1次候補参照動きベクトルを決定するステップと,
前記1次候補参照動きベクトルの,前記復号対象ブロックでの動きベクトル予測における有効性を定量的に表す信頼度を,前記1次候補参照動きベクトルのそれぞれについて復号済みの画像情報を用いて算出するステップと,
前記N個の1次候補参照動きベクトルの中から,信頼度が大きい上位M個(Mは1以上かつN未満の所定の整数)の1次候補参照動きベクトルを,2次候補参照動きベクトルとして選出するステップと,
前記2次候補参照動きベクトルを用いて前記復号対象ブロックの予測動きベクトルを算出し,前記復号した動きベクトル予測残差に前記予測動きベクトルを加算して前記復号対象ブロックの動きベクトルを算出するステップとを有する
動きベクトル予測復号方法。 - 請求項6記載の動きベクトル予測復号方法において,
N個設定すべき前記1次候補参照動きベクトルが前記M個より小さいM′個しか設定できなかった場合に,M′個の1次候補参照動きベクトルを,前記2次候補参照動きベクトルとして選出する
動きベクトル予測復号方法。 - 請求項6または請求項7記載の動きベクトル予測復号方法において,
前記1次候補参照動きベクトルの信頼度を算出するステップでは,前記復号対象ブロックに隣接する復号済み画素の集合をテンプレートとして用い,復号済みの参照画像上で前記1次候補参照動きベクトル分だけ前記テンプレートの領域をずらした領域をマッチング対象領域として設定し,前記テンプレートの前記復号済み画素の集合と前記マッチング対象領域における画素の集合の類似度を前記信頼度として算出する
動きベクトル予測復号方法。 - 請求項6または請求項7記載の動きベクトル予測復号方法において,
前記1次候補参照動きベクトルの信頼度を算出するステップでは,前記1次候補参照動きベクトルに係る1次候補ブロックの動き補償における復号予測残差信号の小ささを信頼度の指標として,前記信頼度を算出する
動きベクトル予測復号方法。 - 請求項6から請求項9までのいずれか1項に記載の動きベクトル予測復号方法において,
前記1次候補参照動きベクトルを決定するステップでは,前記1次候補ブロックの復号に用いた前記動きベクトルと,当該動きベクトルの各々を基準とした所定範囲内の動きベクトルとから,前記1次候補参照動きベクトルを決定する
動きベクトル予測復号方法。 - 符号化対象画像をブロックに分割し,ブロックごとに動き補償を用いて動画像を符号化する動画像符号化装置において,
符号化済みの参照画像を用いて,前記符号化対象画像における符号化対象ブロックの動き探索を行い,動きベクトルを算出する動き探索部と,
前記符号化対象画像における前記符号化対象ブロックの位置に対して,予め定められた相対的な位置にあり,前記符号化対象画像中の符号化済みブロックおよび符号化済み画像における符号化済みブロックの少なくとも一方からなる複数個のブロックを1次候補ブロックとして設定し,前記1次候補ブロックの符号化に用いた動きベクトルからN個(Nは2以上の整数)の1次候補参照動きベクトルを決定する1次候補参照動きベクトル決定部と,
前記1次候補参照動きベクトルの,前記符号化対象ブロックでの動きベクトル予測における有効性を定量的に表す信頼度を,前記1次候補参照動きベクトルのそれぞれについて符号化済みの画像情報を用いて算出する信頼度計算部と,
前記N個の1次候補参照動きベクトルの中から,信頼度が大きい上位M個(Mは1以上かつN未満の所定の整数)の1次候補参照動きベクトルを,2次候補参照動きベクトルとして選出する参照動きベクトル決定部と,
前記2次候補参照動きベクトルを用いて前記符号化対象ブロックの予測動きベクトルを算出し,前記符号化対象ブロックの動き探索で求めた前記動きベクトルと,前記予測動きベクトルとの残差を,前記動きベクトルの符号化情報として符号化する動きベクトル符号化部とを備える
動画像符号化装置。 - ブロックに分割されて符号化された動画像の復号対象画像を,ブロックごとに動き補償を用いて復号する動画像復号装置において,
復号対象ブロックの動きベクトル予測残差を復号する情報源復号部と,
前記復号対象画像における前記復号対象ブロックの位置に対して,予め定められた相対的な位置にあり,前記復号対象画像中の復号済みブロックおよび復号済み画像における復号済みブロックの少なくとも一方からなる複数個のブロックを1次候補ブロックとして設定し,前記1次候補ブロックの復号に用いた動きベクトルからN個(Nは2以上の整数)の1次候補参照動きベクトルを決定する1次候補参照動きベクトル決定部と,
前記1次候補参照動きベクトルの,前記復号対象ブロックでの動きベクトル予測における有効性を定量的に表す信頼度を,前記1次候補参照動きベクトルのそれぞれについて復号済みの画像情報を用いて算出する信頼度計算部と,
前記N個の1次候補参照動きベクトルの中から,信頼度が大きい上位M個(Mは1以上かつN未満の所定の整数)の1次候補参照動きベクトルを,2次候補参照動きベクトルとして選出する参照動きベクトル決定部と,
前記2次候補参照動きベクトルを用いて前記復号対象ブロックの予測動きベクトルを算出し,前記復号した動きベクトル予測残差に前記予測動きベクトルを加算して前記復号対象ブロックの動きベクトルを算出する動きベクトル算出部とを備える
動画像復号装置。 - 請求項1から請求項5までのいずれか1項に記載の動きベクトル予測符号化方法を,コンピュータに実行させるための動きベクトル予測符号化プログラム。
- 請求項6から請求項10までのいずれか1項に記載の動きベクトル予測復号方法を,コンピュータに実行させるための動きベクトル予測復号プログラム。
Priority Applications (9)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
EP11742212.1A EP2536150B1 (en) | 2010-02-09 | 2011-02-08 | Predictive coding method for motion vector, predictive decoding method for motion vector, video coding device, video decoding device, and programs therefor |
CN201180008467.5A CN102823249B (zh) | 2010-02-09 | 2011-02-08 | 运动向量预测编码方法、运动向量预测解码方法、活动图像编码装置、活动图像解码装置 |
BR112012019676A BR112012019676A2 (pt) | 2010-02-09 | 2011-02-08 | método de codificação preditiva de vetor de movimento, método de decodificação preditiva de vetor de movimento, aparelho de codificação de imagem em movimento, aparelho de decodificação de imagem em movimento e programas destes |
US13/576,609 US9497481B2 (en) | 2010-02-09 | 2011-02-08 | Motion vector predictive encoding method, motion vector predictive decoding method, moving picture encoding apparatus, moving picture decoding apparatus, and programs thereof |
RU2012133449/08A RU2519526C2 (ru) | 2010-02-09 | 2011-02-08 | Способ кодирования с предсказанием вектора движения, способ декодирования с предсказанием вектора движения, устройство кодирования фильма, устройство декодирования фильма и их программы |
JP2011553838A JP5367098B2 (ja) | 2010-02-09 | 2011-02-08 | 動きベクトル予測符号化方法,動きベクトル予測復号方法,動画像符号化装置,動画像復号装置およびそれらのプログラム |
ES11742212.1T ES2652337T3 (es) | 2010-02-09 | 2011-02-08 | Procedimiento de codificación predictiva para vector de movimiento, procedimiento de decodificación predictiva para vector de movimiento, dispositivo de codificación de imagen, dispositivo de decodificación de imagen, y programas para ello |
KR1020147015905A KR101528361B1 (ko) | 2010-02-09 | 2011-02-08 | 움직임 벡터 예측 부호화 방법, 움직임 벡터 예측 복호 방법, 동화상 부호화 장치, 동화상 복호 장치 및 그들의 프로그램 |
CA2788954A CA2788954A1 (en) | 2010-02-09 | 2011-02-08 | Motion vector predictive encoding method, motion vector predictive decoding method, moving picture encoding apparatus, moving picture decoding apparatus, and programs thereof |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2010026130 | 2010-02-09 | ||
JP2010-026130 | 2010-02-09 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2011099468A1 true WO2011099468A1 (ja) | 2011-08-18 |
Family
ID=44367741
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/JP2011/052603 WO2011099468A1 (ja) | 2010-02-09 | 2011-02-08 | 動きベクトル予測符号化方法,動きベクトル予測復号方法,動画像符号化装置,動画像復号装置およびそれらのプログラム |
Country Status (11)
Country | Link |
---|---|
US (1) | US9497481B2 (ja) |
EP (1) | EP2536150B1 (ja) |
JP (1) | JP5367098B2 (ja) |
KR (2) | KR20120112724A (ja) |
CN (1) | CN102823249B (ja) |
BR (1) | BR112012019676A2 (ja) |
CA (1) | CA2788954A1 (ja) |
ES (1) | ES2652337T3 (ja) |
RU (1) | RU2519526C2 (ja) |
TW (1) | TWI450592B (ja) |
WO (1) | WO2011099468A1 (ja) |
Cited By (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2012090397A1 (ja) * | 2010-12-28 | 2012-07-05 | 株式会社Jvcケンウッド | 動画像符号化装置、動画像符号化方法及び動画像符号化プログラム、並びに動画像復号装置、動画像復号方法及び動画像復号プログラム |
JP2014531873A (ja) * | 2011-10-04 | 2014-11-27 | クゥアルコム・インコーポレイテッドQualcomm Incorporated | ビデオコーディングのための動きベクトル予測子候補クリッピング削除 |
WO2019187096A1 (ja) * | 2018-03-30 | 2019-10-03 | 株式会社ソシオネクスト | 復号方法、復号装置、符号化装置及びプログラム |
Families Citing this family (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2009032255A2 (en) * | 2007-09-04 | 2009-03-12 | The Regents Of The University Of California | Hierarchical motion vector processing method, software and devices |
CA2788946A1 (en) | 2010-02-09 | 2011-08-18 | Nippon Telegraph And Telephone Corporation | Motion vector predictive encoding method, motion vector predictive decoding method, moving picture encoding apparatus, moving picture decoding apparatus, and programs thereof |
CA2788952A1 (en) * | 2010-02-09 | 2011-08-18 | Nippon Telegraph And Telephone Corporation | Motion vector predictive encoding method, motion vector predictive decoding method, moving picture encoding apparatus, moving picture decoding apparatus, and programs thereof |
CN102884793B (zh) * | 2010-02-09 | 2016-03-23 | 日本电信电话株式会社 | 运动向量预测编码方法、运动向量预测解码方法、活动图像编码装置及活动图像解码装置 |
CN103445152A (zh) * | 2013-01-05 | 2013-12-18 | 胡素芳 | 一种萝卜汁加工方法 |
CN103248899B (zh) * | 2013-05-21 | 2016-02-24 | 清华大学 | 多视点视频skip模式的运动矢量预测方法 |
JP6577852B2 (ja) * | 2015-12-03 | 2019-09-18 | キヤノン株式会社 | 動きベクトル検出装置およびその制御方法 |
US10834419B2 (en) * | 2016-04-13 | 2020-11-10 | Qualcomm Incorporated | Conformance constraint for collocated reference index in video coding |
RU2741507C1 (ru) | 2017-10-09 | 2021-01-26 | Нокиа Текнолоджиз Ой | Устройство и способ для кодирования и декодирования видео |
EP3780608A4 (en) | 2018-04-02 | 2021-12-01 | SZ DJI Technology Co., Ltd. | IMAGE PROCESSING PROCESS AND IMAGE PROCESSING DEVICE |
WO2019234674A1 (en) | 2018-06-07 | 2019-12-12 | Beijing Bytedance Network Technology Co., Ltd. | Integer motion compensation |
TWI719519B (zh) | 2018-07-02 | 2021-02-21 | 大陸商北京字節跳動網絡技術有限公司 | 對於dmvr的塊尺寸限制 |
CN118337962B (zh) * | 2024-06-12 | 2024-09-03 | 湖南中泓汇智智能科技有限公司 | 一种用于超视距远程驾驶平台的5g网络数据传输方法 |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2008171247A (ja) * | 2007-01-12 | 2008-07-24 | Sony Corp | 動き検出装置、方法、及びプログラム |
JP2008283490A (ja) * | 2007-05-10 | 2008-11-20 | Ntt Docomo Inc | 動画像符号化装置、方法及びプログラム、並びに動画像復号化装置、方法及びプログラム |
WO2010001917A1 (ja) * | 2008-07-01 | 2010-01-07 | ソニー株式会社 | 画像処理装置および方法 |
JP2010026130A (ja) | 2008-07-17 | 2010-02-04 | Ricoh Co Ltd | 画像形成装置、画像形成方法およびプリント媒体 |
Family Cites Families (39)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP3263807B2 (ja) | 1996-09-09 | 2002-03-11 | ソニー株式会社 | 画像符号化装置および画像符号化方法 |
FR2756399B1 (fr) | 1996-11-28 | 1999-06-25 | Thomson Multimedia Sa | Procede et dispositif de compression video pour images de synthese |
WO1998042134A1 (en) | 1997-03-17 | 1998-09-24 | Mitsubishi Denki Kabushiki Kaisha | Image encoder, image decoder, image encoding method, image decoding method and image encoding/decoding system |
US6011870A (en) | 1997-07-18 | 2000-01-04 | Jeng; Fure-Ching | Multiple stage and low-complexity motion estimation for interframe video coding |
JP4573366B2 (ja) | 1997-09-25 | 2010-11-04 | 株式会社大宇エレクトロニクス | 動きベクトル符号化方法及び符号化装置 |
JP3540142B2 (ja) | 1998-01-30 | 2004-07-07 | 株式会社東芝 | 動きベクトル検出回路および動きベクトル検出方法 |
US6058143A (en) * | 1998-02-20 | 2000-05-02 | Thomson Licensing S.A. | Motion vector extrapolation for transcoding video sequences |
ES2665693T3 (es) | 2001-11-06 | 2018-04-26 | Panasonic Intellectual Property Corporation Of America | Método de codificación de imágenes en movimiento y método de decodificación de imágenes en movimiento |
JP2003224854A (ja) | 2002-01-29 | 2003-08-08 | Hitachi Ltd | 動きベクトル検出装置及び画像処理装置並びにコンピュータ・ソフトウエア |
RU2314656C2 (ru) | 2002-06-11 | 2008-01-10 | Нокиа Корпорейшн | Внутреннее кодирование, основанное на пространственном прогнозировании |
JP2004023458A (ja) * | 2002-06-17 | 2004-01-22 | Toshiba Corp | 動画像符号化/復号化方法及び装置 |
JP4724351B2 (ja) * | 2002-07-15 | 2011-07-13 | 三菱電機株式会社 | 画像符号化装置、画像符号化方法、画像復号装置、画像復号方法、および通信装置 |
KR100865034B1 (ko) | 2002-07-18 | 2008-10-23 | 엘지전자 주식회사 | 모션 벡터 예측 방법 |
RU2338332C2 (ru) | 2002-11-21 | 2008-11-10 | Эл Джи Электроникс Инк. | Способ предсказания блока изображения с использованием усовершенствованного режима прямого предсказания |
JP4003128B2 (ja) * | 2002-12-24 | 2007-11-07 | ソニー株式会社 | 画像データ処理装置および方法、記録媒体、並びにプログラム |
US7336707B2 (en) * | 2003-06-06 | 2008-02-26 | Samsung Electronics Co., Ltd. | Method and apparatus for detecting improper area for motion compensation in video signal |
KR20050078706A (ko) | 2004-01-31 | 2005-08-08 | 삼성전자주식회사 | 메모리 액세스 방법 및 메모리 액세스 장치 |
JP4937741B2 (ja) | 2005-01-07 | 2012-05-23 | 日本電信電話株式会社 | 映像符号化方法及び装置、映像復号方法及び装置、それらのプログラムおよびそれらプログラムを記録した記録媒体 |
US20060153300A1 (en) | 2005-01-12 | 2006-07-13 | Nokia Corporation | Method and system for motion vector prediction in scalable video coding |
BRPI0610398B1 (pt) | 2005-04-13 | 2019-07-02 | Nokia Technologies Oy | Método e aparelho |
KR101276720B1 (ko) | 2005-09-29 | 2013-06-19 | 삼성전자주식회사 | 카메라 파라미터를 이용하여 시차 벡터를 예측하는 방법,그 방법을 이용하여 다시점 영상을 부호화 및 복호화하는장치 및 이를 수행하기 위한 프로그램이 기록된 기록 매체 |
JP2008109632A (ja) * | 2006-09-28 | 2008-05-08 | Toshiba Corp | 動きベクトル検出装置及びその方法 |
KR101083379B1 (ko) | 2007-03-14 | 2011-11-14 | 니폰덴신뎅와 가부시키가이샤 | 움직임 벡터 탐색 방법 및 장치 그리고 프로그램을 기록한 기록매체 |
WO2008111005A1 (en) | 2007-03-15 | 2008-09-18 | Nokia Corporation | System and method for providing improved residual prediction for spatial scalability in video coding |
RU2432703C2 (ru) | 2007-04-17 | 2011-10-27 | Нокиа Корпорейшн | Масштабируемое видеокодирование с обратной связью |
TW200905030A (en) | 2007-07-31 | 2009-02-01 | Chinese Professional Fabric Industry Co Ltd | A manufacturing method of optical-excitation fabric |
TW200910971A (en) | 2007-08-22 | 2009-03-01 | Univ Nat Cheng Kung | Direction detection algorithms for H.264 intra prediction |
KR101228020B1 (ko) * | 2007-12-05 | 2013-01-30 | 삼성전자주식회사 | 사이드 매칭을 이용한 영상의 부호화 방법 및 장치, 그복호화 방법 및 장치 |
US8660175B2 (en) * | 2007-12-10 | 2014-02-25 | Qualcomm Incorporated | Selective display of interpolated or extrapolated video units |
US8165210B2 (en) | 2007-12-17 | 2012-04-24 | Vixs Systems, Inc. | Video codec with shared interpolation filter and method for use therewith |
JP4513034B2 (ja) * | 2008-02-20 | 2010-07-28 | ソニー株式会社 | 画像信号処理装置、画像信号処理方法、およびプログラム |
KR20090094595A (ko) | 2008-03-03 | 2009-09-08 | 삼성전자주식회사 | 복수 참조에 의한 움직임 예측을 이용한 부호화 방법 및장치, 그리고 복수 참조에 의한 움직임 예측을 이용한복호화 방법 및 장치 |
JP2011515060A (ja) * | 2008-03-09 | 2011-05-12 | エルジー エレクトロニクス インコーポレイティド | ビデオ信号のエンコーディングまたはデコーディング方法及び装置 |
EP2266318B1 (en) | 2008-03-19 | 2020-04-22 | Nokia Technologies Oy | Combined motion vector and reference index prediction for video coding |
KR101364195B1 (ko) | 2008-06-26 | 2014-02-21 | 에스케이텔레콤 주식회사 | 움직임벡터 부호화/복호화 방법 및 그 장치 |
WO2010041855A2 (en) * | 2008-10-06 | 2010-04-15 | Lg Electronics Inc. | A method and an apparatus for processing a video signal |
CA2788952A1 (en) * | 2010-02-09 | 2011-08-18 | Nippon Telegraph And Telephone Corporation | Motion vector predictive encoding method, motion vector predictive decoding method, moving picture encoding apparatus, moving picture decoding apparatus, and programs thereof |
CA2788946A1 (en) * | 2010-02-09 | 2011-08-18 | Nippon Telegraph And Telephone Corporation | Motion vector predictive encoding method, motion vector predictive decoding method, moving picture encoding apparatus, moving picture decoding apparatus, and programs thereof |
CN102884793B (zh) * | 2010-02-09 | 2016-03-23 | 日本电信电话株式会社 | 运动向量预测编码方法、运动向量预测解码方法、活动图像编码装置及活动图像解码装置 |
-
2011
- 2011-02-08 JP JP2011553838A patent/JP5367098B2/ja active Active
- 2011-02-08 US US13/576,609 patent/US9497481B2/en active Active
- 2011-02-08 EP EP11742212.1A patent/EP2536150B1/en active Active
- 2011-02-08 CN CN201180008467.5A patent/CN102823249B/zh active Active
- 2011-02-08 ES ES11742212.1T patent/ES2652337T3/es active Active
- 2011-02-08 KR KR1020127020437A patent/KR20120112724A/ko active Application Filing
- 2011-02-08 BR BR112012019676A patent/BR112012019676A2/pt not_active IP Right Cessation
- 2011-02-08 TW TW100104093A patent/TWI450592B/zh active
- 2011-02-08 CA CA2788954A patent/CA2788954A1/en not_active Abandoned
- 2011-02-08 RU RU2012133449/08A patent/RU2519526C2/ru active
- 2011-02-08 WO PCT/JP2011/052603 patent/WO2011099468A1/ja active Application Filing
- 2011-02-08 KR KR1020147015905A patent/KR101528361B1/ko active IP Right Grant
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2008171247A (ja) * | 2007-01-12 | 2008-07-24 | Sony Corp | 動き検出装置、方法、及びプログラム |
JP2008283490A (ja) * | 2007-05-10 | 2008-11-20 | Ntt Docomo Inc | 動画像符号化装置、方法及びプログラム、並びに動画像復号化装置、方法及びプログラム |
WO2010001917A1 (ja) * | 2008-07-01 | 2010-01-07 | ソニー株式会社 | 画像処理装置および方法 |
JP2010026130A (ja) | 2008-07-17 | 2010-02-04 | Ricoh Co Ltd | 画像形成装置、画像形成方法およびプリント媒体 |
Non-Patent Citations (4)
Title |
---|
KADONO; KIKUCHI; SUZUKI: "H.264/AVC textbook", 2009, IMPRESS R&D, pages: 123 - 125 |
KOBAYASHI; SUZUKI; BOON; HORIKOSHI: "Reduction of Predictive Information Amount with Motion Prediction Method Using Template Matching", THE PROCEEDINGS OF PICTURE CODING SYMPOSIUM OF JAPAN, 2005, pages 17 - 18 |
See also references of EP2536150A4 |
T. YAMAMOTO: "A new scheme for motion vector predictor encoding", ITU-T SG16/Q6, April 2007 (2007-04-01) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO2012090397A1 (ja) * | 2010-12-28 | 2012-07-05 | 株式会社Jvcケンウッド | 動画像符号化装置、動画像符号化方法及び動画像符号化プログラム、並びに動画像復号装置、動画像復号方法及び動画像復号プログラム |
JP2014531873A (ja) * | 2011-10-04 | 2014-11-27 | クゥアルコム・インコーポレイテッドQualcomm Incorporated | ビデオコーディングのための動きベクトル予測子候補クリッピング削除 |
US9083983B2 (en) | 2011-10-04 | 2015-07-14 | Qualcomm Incorporated | Motion vector predictor candidate clipping removal for video coding |
WO2019187096A1 (ja) * | 2018-03-30 | 2019-10-03 | 株式会社ソシオネクスト | 復号方法、復号装置、符号化装置及びプログラム |
JPWO2019187096A1 (ja) * | 2018-03-30 | 2021-04-08 | 株式会社ソシオネクスト | 復号方法、復号装置、符号化装置及びプログラム |
US11197011B2 (en) | 2018-03-30 | 2021-12-07 | Socionext Inc. | Decoding method |
JP7248013B2 (ja) | 2018-03-30 | 2023-03-29 | 株式会社ソシオネクスト | 復号方法、復号装置、符号化装置及びプログラム |
Also Published As
Publication number | Publication date |
---|---|
JPWO2011099468A1 (ja) | 2013-06-13 |
EP2536150B1 (en) | 2017-09-13 |
ES2652337T3 (es) | 2018-02-01 |
EP2536150A1 (en) | 2012-12-19 |
CA2788954A1 (en) | 2011-08-18 |
KR20120112724A (ko) | 2012-10-11 |
US20120294371A1 (en) | 2012-11-22 |
US9497481B2 (en) | 2016-11-15 |
JP5367098B2 (ja) | 2013-12-11 |
RU2519526C2 (ru) | 2014-06-10 |
CN102823249A (zh) | 2012-12-12 |
RU2012133449A (ru) | 2014-03-27 |
KR101528361B1 (ko) | 2015-06-12 |
TW201210349A (en) | 2012-03-01 |
EP2536150A4 (en) | 2014-06-04 |
TWI450592B (zh) | 2014-08-21 |
CN102823249B (zh) | 2016-07-06 |
KR20140092904A (ko) | 2014-07-24 |
BR112012019676A2 (pt) | 2016-05-03 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
JP5367098B2 (ja) | 動きベクトル予測符号化方法,動きベクトル予測復号方法,動画像符号化装置,動画像復号装置およびそれらのプログラム | |
JP5306486B2 (ja) | 動きベクトル予測符号化方法,動きベクトル予測復号方法,動画像符号化装置,動画像復号装置およびそれらのプログラム | |
JP5367097B2 (ja) | 動きベクトル予測符号化方法、動きベクトル予測復号方法、動画像符号化装置、動画像復号装置およびそれらのプログラム | |
JP5306485B2 (ja) | 動きベクトル予測符号化方法、動きベクトル予測復号方法、動画像符号化装置、動画像復号装置およびそれらのプログラム | |
JP5216710B2 (ja) | 復号化処理方法 | |
JP5281597B2 (ja) | 動きベクトル予測方法,動きベクトル予測装置および動きベクトル予測プログラム | |
JP5281596B2 (ja) | 動きベクトル予測方法,動きベクトル予測装置および動きベクトル予測プログラム |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
WWE | Wipo information: entry into national phase |
Ref document number: 201180008467.5 Country of ref document: CN |
|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 11742212 Country of ref document: EP Kind code of ref document: A1 |
|
DPE1 | Request for preliminary examination filed after expiration of 19th month from priority date (pct application filed from 20040101) | ||
WWE | Wipo information: entry into national phase |
Ref document number: 2011553838 Country of ref document: JP |
|
ENP | Entry into the national phase |
Ref document number: 2788954 Country of ref document: CA |
|
WWE | Wipo information: entry into national phase |
Ref document number: 13576609 Country of ref document: US |
|
ENP | Entry into the national phase |
Ref document number: 20127020437 Country of ref document: KR Kind code of ref document: A |
|
REEP | Request for entry into the european phase |
Ref document number: 2011742212 Country of ref document: EP |
|
WWE | Wipo information: entry into national phase |
Ref document number: 6853/CHENP/2012 Country of ref document: IN Ref document number: 2011742212 Country of ref document: EP |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
WWE | Wipo information: entry into national phase |
Ref document number: 2012133449 Country of ref document: RU |
|
REG | Reference to national code |
Ref country code: BR Ref legal event code: B01A Ref document number: 112012019676 Country of ref document: BR |
|
ENP | Entry into the national phase |
Ref document number: 112012019676 Country of ref document: BR Kind code of ref document: A2 Effective date: 20120806 |