WO2014200313A1 - 움직임 정보를 획득하기 위한 방법 - Google Patents
움직임 정보를 획득하기 위한 방법 Download PDFInfo
- Publication number
- WO2014200313A1 WO2014200313A1 PCT/KR2014/005226 KR2014005226W WO2014200313A1 WO 2014200313 A1 WO2014200313 A1 WO 2014200313A1 KR 2014005226 W KR2014005226 W KR 2014005226W WO 2014200313 A1 WO2014200313 A1 WO 2014200313A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- block
- motion vector
- current
- subblock
- predictor
- Prior art date
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/50—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
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/134—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the element, parameter or criterion affecting or controlling the adaptive coding
- H04N19/136—Incoming video signal characteristics or properties
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/10—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding
- H04N19/169—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding
- H04N19/17—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object
- H04N19/176—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using adaptive coding characterised by the coding unit, i.e. the structural portion or semantic portion of the video signal being the object or the subject of the adaptive coding the unit being an image region, e.g. an object the region being a block, e.g. a macroblock
-
- 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
- H04N19/44—Decoders specially adapted therefor, e.g. video decoders which are asymmetric with respect to the encoder
-
- 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
Definitions
- the present invention relates to a method for obtaining motion information in video encoding or decoding.
- Predictive coding includes inter-frame prediction coding and intra-frame prediction coding, which remove the correlation of video signals in the time axis and the video signals in the spatial axis, respectively. ).
- the inter-frame prediction encoding may predict an image of the currently encoded frame according to the image of the encoded frame. Prediction residuals are transformed and entropy coded.
- the first encoded frame for predicting the current encoded frame is called a reference frame.
- the spatial positional relationship between the prediction block on the reference frame and the currently encoded block may be decoded into a motion vector.
- the encoder obtains a motion vector through motion estimation, predicts the motion vector, and records the motion vector difference between the motion vector and the predictor of the motion vector in a bit-stream.
- the decoder predicts the motion vector in the same way and reconstructs the motion vector according to the difference between the predictor and the motion vector of the motion vector read from the bit stream.
- the encoder / decoder derives the motion vector directly in the same way. The second approach eliminates the need to send motion vector differences within the bit stream.
- motion information is derived.
- the derived motion information may be a predictor of the motion vector, a motion vector, a candidate of the predictor of the motion vector, or a candidate of the motion vector.
- the motion vector is predicted by the median prediction method in the inter-frame prediction mode.
- the median prediction method three coded blocks adjacent to the current coded block are selected, the median of the horizontal components of the motion vectors of the three coded blocks and the median of the vertical components of the motion vectors of the three coded blocks. Are respectively adopted as horizontal and vertical components of the predictor of the motion vector of the currently encoded block.
- temporal direct mode the motion vector of a block of a reference frame having the same spatial position as the current coded block is scaled and adopted as the motion vector of the current coded block.
- motion vectors are predicted by the distance median method in inter-frame prediction mode.
- the distance median method three coded blocks close to the current coded block in space are selected, and a distance between two motion vectors among the motion vectors of the three coded blocks is calculated and corresponds to the median of the calculated distances.
- the motion vector is adopted as a predictor of the motion vector of the currently encoded block.
- the temporal direct mode the motion vector of the block of the reference frame having the same spatial position as the current coded block is scaled and adopted as the motion vector of the currently coded block.
- a motion vector competition scheme In the latest international video coding standard, High Efficiency Video Coding (HEVC), a motion vector competition scheme has been introduced.
- a candidate motion vector set is derived according to a rule, and a suitable candidate motion vector is selected from the candidate motion vector sets, and is adopted as a predictor of the currently encoded motion vector.
- the selected motion vector is transmitted in the bit stream.
- Candidate motion vector set may include spatial candidate motion vectors and temporal candidate motion vectors.
- the spatial candidate motion vectors will be selected according to a specific rule and priority from the motion vectors of the left and top blocks adjacent to the current coded block.
- Temporal candidate motion vectors may be selected according to a specific rule and priority from motion vectors of blocks of a reference frame having a spatial position adjacent to or the same as the spatially encoded block.
- the motion vector is derived or predicted only once, and the derived or predicted value is adopted as the final motion vector or predictor.
- motion vector differences of blocks that are spatially or temporally adjacent to each other still have correlations, and correlations between motion vector differences are not used in a conventional video encoding or decoding method.
- a motion vector prediction or derivation method based on a median or distance only motion vectors of blocks located in the same frame of the current coded block are taken into account, so that only correlations of motion vectors in the spatial domain are to be removed. Can be.
- the motion vector differences still correlate with the motion vectors of the time domain.
- the motion vector derivation method In the temporal direct mode of H.264 / AVC and AVS, the motion vector derivation method only considers motion vectors of blocks on reference frames located at the same spatial location of the current coded block.
- the motion vector derived through this method utilizes a temporal correlation but does not consider the correlation between the current coded block in the same frame and blocks adjacent to the current coded block. Thus, the derived motion vectors are not optimal.
- the motion vector prediction and derivation method based on the competition scheme of HEVC even if both temporal and spatial candidate motion vectors are included in the candidate motion vector set and the optimal motion vector is selected, the motion vector differences of adjacent blocks still have correlations. have.
- An embodiment of the present invention provides a method for acquiring motion information in video encoding or decoding including a method of determining motion information of a current block or a subblock of a video image.
- Another embodiment of the present invention provides an apparatus for obtaining motion information in video encoding or decoding.
- a method for acquiring motion information in video encoding or decoding includes determining motion information of a current block or subblock of a video image. This process
- An apparatus for obtaining motion information in video encoding or decoding
- a calculator configured to determine a second predictor of the motion vector of the current block or the subblock according to the first predictor of the motion vector of the current block or the subblock.
- the first reference block set first includes reference blocks for predicting or deriving the motion vector of the current block or sub-block,
- the second reference block set may include reference blocks for predicting the first motion vector difference of the current block or subblock.
- the motion vector is derived or predicted only once.
- the motion vector difference of spatially or temporally adjacent blocks still has a correlation. Therefore, in order to fully utilize the correlation of motion vector differences, the present invention can derive or predict the motion vector more accurately by deriving or predicting the motion vector twice.
- FIG. 1 is a flowchart illustrating a method of obtaining a motion vector according to an embodiment of the present invention.
- FIG. 2 is a flowchart of a first embodiment according to the present invention.
- FIG. 3 is a schematic diagram illustrating a reference block of a first category of a current block according to an embodiment of the present invention.
- FIG. 4 is a flowchart of a second embodiment according to the present invention.
- FIG. 5 is a schematic diagram illustrating a reference block of a second category of a current block according to an embodiment of the present invention.
- FIG. 6 is a flowchart of a third embodiment according to the present invention.
- FIG. 7 is a flowchart of a fourth embodiment according to the present invention.
- FIG. 8 is a flowchart of a fifth embodiment according to the present invention.
- FIG. 9 is a schematic diagram illustrating a reference block of a first category of a current block according to another embodiment of the present invention.
- FIG. 10 is a schematic diagram illustrating a reference block of a second category of a current block according to another embodiment of the present invention.
- FIG. 11 is a flowchart of a sixth embodiment according to the present invention.
- FIG. 12 is a schematic diagram illustrating a reference block of a first category of a current block according to another embodiment of the present invention.
- FIG. 13 is a schematic diagram illustrating a reference block of a second category of a current block according to another embodiment of the present invention.
- FIG. 14 is a schematic diagram illustrating an apparatus for obtaining motion information according to an embodiment of the present invention.
- 15 is a schematic diagram showing a configuration of an input unit in an apparatus according to an embodiment of the present invention.
- 16 is a schematic diagram showing a configuration of a predictor in an apparatus according to an embodiment of the present invention.
- the present invention provides a method and apparatus for obtaining motion information in video encoding or decoding. Encoding performance can be improved with the method and apparatus.
- the solution of the present invention is implemented as follows.
- a method for acquiring motion information in video encoding or decoding includes determining motion information of a current block or subblock of a video image. This process
- the first reference block set first includes reference blocks for predicting or deriving a motion vector of the current block or subblock,
- the second reference block set may include reference blocks for predicting a first motion vector difference of the current block or subblock.
- the first reference block set of the current block includes an encoded block or a decoded block adjacent to the current block in a current frame, or an encoded block that is spatially adjacent to or overlaps with the current block in a reference frame of the current frame, or Contains a decoded block,
- the second reference block set of the current block includes a coded block or a decoded block in the current frame adjacent to the current block, or a coded block that is spatially adjacent to or overlaps with the current block within the reference frame of the current frame.
- the first reference block set of the current subblock includes a coded block or a decoded block in the current frame adjacent to the current subblock or the block of the current subblock, or spatially the current subblock or the current block.
- the second reference block set of the current subblock includes a coded block or a decoded block in the current frame adjacent to the current subblock or the block of the current subblock, or spatially the current subblock or the current block. It may include a block encoded or decoded in the reference frame of the current frame adjacent or overlapping the block of the sub-block.
- the first predictor of the motion vector of the current block or subblock according to the motion vector of each reference block in the first reference block set of the current block or subblock, Recording the derivation information of the first predictor of the motion vector of the current block or sub-block, and
- the first predictor of the motion vector of the current block or subblock and the current in the bit stream according to the motion vector of each reference block in the first reference block set of the current block or subblock And deriving the derivation information of the first predictor of the motion vector of the block or subblock.
- the first motion vector difference between the motion vector of each reference block in the second reference block set and the first predictor of the motion vector of the reference block in the second reference block set of the current block or subblock In the determining step,
- the motion information of the current block or sub-block When the motion information of the current block or sub-block is determined, obtain the motion vector of each reference block in the second reference block set and the first predictor of the motion vector of the reference block in the second reference block set and The motion vector of the reference block in the second reference block set and the reference in the second reference block set with a difference of the first motion vector of the reference block of the second reference block set of the current block or sub-block Selecting a difference between the first predictors of the motion vectors of the block,
- the method for obtaining the motion information is the first motion of the current block or subblock. Setting the predictor of vector difference to a zero vector, or
- the method for obtaining the motion information is the first motion vector of the current block or subblock. And setting the predictor of the difference to a zero vector.
- the predictor of the first motion vector difference of the current block or subblock according to the first motion vector difference of each reference block of the second reference block set, Recording the derivation information of the predictor of one motion vector difference into a bit stream, and by the decoder, the decoder of the current block or subblock according to the first motion vector difference of each reference block in the second reference block set; And deriving the derivation information of the predictor of the first motion vector difference and the predictor of the first motion vector difference of the current block or subblock.
- the sum of the predictor of the first motion vector difference of the current block or subblock and the first predictor of the motion vector of the current block or subblock or of the first motion vector difference of the current block or subblock And adopting the weighted sum of the predictor and the first predictor of the motion vector of the current block or subblock as the second predictor of the motion vector of the current block or subblock.
- the reference blocks in the first reference block set of the current block or sub block may be the same as or different from the reference blocks in the second reference block set of the current block or sub block.
- the reference blocks in the first reference block set of the subblock are the same as or different from the reference blocks in the first reference block set of another subblock, and in the second reference block set of the subblock.
- the reference blocks may be the same as or different from the reference blocks in the second reference block set of another sub block.
- the reference block may be a sub block.
- An apparatus for obtaining motion information in video encoding or decoding
- a calculator configured to determine a second predictor of the motion vector of the current block or the subblock according to the first predictor of the motion vector of the current block or subblock,
- the first reference block set first includes reference blocks for predicting or deriving the motion vector of the current block or sub-block,
- the second reference block set may include reference blocks for predicting the first motion vector difference of the current block or subblock.
- the motion vector of the current block or subblock is predicted to obtain a first predictor of the motion vector of the current block or subblock.
- the first motion vector difference of the current block or subblock is the first motion vector difference of the reference blocks in the second reference block set to derive the predictor of the first motion vector difference of the current block or subblock.
- the second predictor of the motion vector of the current block or subblock is the first prediction of the motion vector of the current block or subblock with the predictor of the first motion vector difference of the current block or subblock.
- the final predictor of the motion vector of the current block Acquired according to one predictor, the final predictor of the motion vector of the current block, the candidate of the predictor of the motion vector of the current block, the motion vector of the current block, or the candidate of the motion vector of the current block.
- the correlation between the motion vector differences of the reference block and the current block or subblock can be fully utilized to estimate the difference of the motion vector of the current block / subblock, and also the current block or subblock
- the first predictor of the motion vector of the current block or subblock may be corrected to obtain an accurate second predictor of the motion vector of.
- the accuracy of motion information and encoding performance can be greatly improved.
- the motion vector is derived or predicted only once.
- the motion vector difference of spatially or temporally adjacent blocks still has a correlation. Therefore, encoding performance can be improved by utilizing the correlation.
- the present invention provides a method for obtaining a motion vector as shown in FIG.
- the first predictor of the motion vector of the current block or subblock is calculated according to the motion vector of each reference block in the first reference block set of the current block or subblock.
- the first reference block set first includes reference blocks to predict or derive a motion vector of the current block or subblock.
- Vector differences are determined.
- the first motion vector difference between the motion vector of the current block or subblock and the first predictor of the motion vector of the current block or subblock is referred to as a second reference to obtain a predictor of the first motion vector difference of the current block or subblock. It is predicted according to the first motion vector difference of each reference block in the block set.
- the second reference block set includes reference blocks to predict the first motion vector difference of the current block or subblock.
- the second predictor of the motion vector of the current block or subblock is determined according to the predictor of the first motion vector difference of the current block or subblock and the first predictor of the motion vector of the current block or subblock. .
- the second predictor of the motion vector obtained in the above flowchart is a predictor of a motion vector of the current block or subblock, a candidate of a motion vector of the current block or subblock, which may be used for encoding or decoding the current block or subblock. It may be adopted as a candidate of the motion vector of the current block or subblock or the motion vector of the current block or subblock.
- a method for obtaining motion information can be regarded as an auxiliary prediction or derivation method of a motion vector. In the prediction process of the motion vector, the motion vector difference initially generated after the prediction of the motion vector may be predicted again to obtain a predictor of the motion vector difference.
- the last predictor of the motion vector can be obtained according to the result of the two predictions.
- the derived value of the motion vector is first obtained, and the derived value of the motion vector difference is obtained according to the information of the first motion vector differences of the adjacent blocks.
- the last derived variable of the motion vector is obtained according to the two derived values.
- the derived value of the motion vector of the current block or subblock initially obtained through the prediction process and the motion vector of the current block or subblock obtained through the derivation process are referred to as the first predictor of the motion vector.
- the estimation result of the obtained motion vector resulting from two predictions or two derivatives is referred to as a second predictor of the motion vector.
- the method of transmitting the derived information in the bit stream may be applied to the calculation of the first predictor of the motion vector of the current block or subblock and / or the predictor of the first motion vector difference of the current block or subblock.
- the encoder records the derivation information of the first predictor of the motion vector and / or the derivation information of the predictor of the first motion vector difference in the bit stream.
- the decoder then calculates the first predictor of the motion vector and / or the predictor of the first motion vector difference according to the derivation information of the bit stream.
- the blocks of the current frame and a reference frame that are spatially close to or overlapping with the current block are both used in the two prediction or derivation processes of the present invention, or respectively in the two prediction or derivation processes of the present invention. Can be used. Therefore, the spatial and temporal correlations of the motion vectors of the other blocks can be eliminated to the maximum, and the coding efficiency of the motion vectors can be increased.
- a method for obtaining motion information is applied to each sub block to obtain motion information of each sub block.
- a method for acquiring motion information in a block and a subblock may be implemented according to embodiments of the present invention. The solution of the present invention is described with reference to the embodiments.
- motion information is determined for the current block. And a flowchart for deriving motion information of the current block is provided in each embodiment.
- the current block is divided into several sub blocks. The motion information is determined for each sub block.
- a flowchart for deriving motion information of sub-blocks is provided in each embodiment.
- the first embodiment is described as follows.
- FIG. 2 is a flowchart of a first embodiment according to the present invention.
- 3 is a schematic diagram illustrating reference blocks of a first category of a current block according to an embodiment of the present invention. As shown in FIG. 2, the method for acquiring motion information according to the first embodiment includes the following blocks.
- the first reference block set and the second reference block set include at least one of a reference block of the first category and a reference block of the first category.
- Reference blocks of the first category are encoded or decoded blocks of the current frame adjacent to the current block.
- Reference blocks of the second category are encoded or decoded blocks of reference frames that spatially overlap or adjacent to current blocks.
- block 201 may be implemented as follows.
- the first reference block set only includes reference blocks of the first category.
- the reference blocks of the first category of the current block include AL1, AL2, AU1, AU2, AU3, AC1 and AC2.
- the reference blocks in the first reference block set of the current block are adjacent block AU1 located above the upper left corner of the current block, adjacent block AL2 located to the left of the upper left corner of the current block, and upper left corner of the upper left corner of the current block. It can be configured as an adjacent block AC2 located at.
- the first predictor of the motion vector of the current block is calculated according to the motion vector of each reference block in the first reference block set obtained in block 202.
- the first predictor of the motion vector of the current block may be calculated according to the same method for predicting and deriving the motion vector of the current block used in the existing encoding or decoding method.
- Block 203 may be implemented as follows.
- the average value of the horizontal components of the motion vectors obtained in block 202 and the average value of the vertical components of the motion vectors obtained in block 202 are respectively calculated, and the horizontal component of the predictor of the motion vector of the current block and the current block Adopt as the vertical component of the predictor of the motion vector. If the motion vector of the specific reference block in the first reference block set does not exist, the motion vector that does not exist is set to a zero vector.
- the motion vector of the reference block in the first reference block set is determined in block 203 before the first predictor of the motion vector of the current block is determined. It may be scaled according to the distance between the frame in which the frame is located and the reference frame of the frame in which the reference block is located in the first reference block set and the distance between the frame in which the current block is located and the reference frame of the frame in which the current block is located.
- the first predictor of the motion vector of the current block may be calculated according to the scaled motion vector. By doing so, the accuracy of the first predictor can be improved.
- a second reference block set of the current block is determined.
- Block 204 can be implemented as follows.
- the second reference block set only includes reference blocks of the first category.
- the reference blocks of the first category of the current block include AL1, AL2, AU1, AU2, AU3, AC1 and AC2.
- the reference blocks in the second reference block set of the current block are adjacent block AU1 located above the upper left corner of the current block, adjacent block AL2 located to the left of the upper left corner of the current block, and upper left corner of the upper left corner of the current block. It can be configured as an adjacent block AC2 located at.
- a motion vector of each reference block in the second reference block set and a first predictor of the motion vector of each reference block in the second reference block set are obtained.
- the first motion vector difference between the motion vector of the reference block and the first predictor of the motion vector of the reference block is calculated.
- the reference blocks in the second reference block set are all coded blocks. Therefore, the first predictors of the motion vectors of the reference blocks in the second reference block set and the motion vectors of the reference blocks in the second reference block set are generated.
- the motion vector of each reference block and the first predictor of the motion vector of the reference block are obtained directly.
- the first predictor of the motion vector of the reference block is subtracted from the motion vector of the reference block.
- the distance between the frame in which the reference block in the second reference block set is located and the reference frame of the frame in which the reference block in the second reference block set is located is a reference of the frame in which the current block is located and the frame in which the current block is located. It may be different from the distance between the frames. Accordingly, in order to improve encoding or decoding performance, in block 205, the first predictor of the motion vector of the reference block in the second reference block set and the motion vector of the reference block in the second reference block set is determined by the second reference block set.
- the distance between the frame in which the reference block in the second reference block set is located and the reference frame in the frame in which the reference block in the second reference block set is located and the frame in which the current block is located and the current block before the first motion vector difference of the reference block in the block is calculated It may be scaled according to the distance between reference frames of the located frame. By doing so, the accuracy of the first motion vector difference of the reference blocks in the second reference block set can be improved.
- the first predictor of the motion vector of reference block A is determined by the distance between the frame in which the reference block in the first reference block set is located and the reference frame of the frame in which the reference block in the first reference block set is located. It may be scaled according to a distance between a frame located and a reference frame of the frame in which the current block is located. Or, when the reference block A in the second reference block set is determined as the motion information of the current block and the reference block A is derived with reference to FIG.
- the motion of the reference block B in the first reference block set of the reference block A is determined. If the vector is not fully scaled at block 203 (i.e., the scale ratio) is the distance between the frame where reference block A is located and the reference frame of the frame where reference block A is located and the frame where reference block B is located and reference block B ), The first predictor of the motion vector of the reference block A may be scaled to completely compensate for the distance difference. And the scaling process of the above paragraph and the calculation of the first motion vector difference of each reference block in the second reference block set can be applied to the scaled motion vectors.
- the predictor of the first motion vector difference of the current block at block 206 is calculated according to the first motion vector difference of each reference block in the second set of reference blocks calculated at block 205.
- the first motion vector difference of the current block is predicted according to the first motion vector difference of the reference block in the second reference block set.
- the prediction method may be selected according to the requirements. For example, the weighted average of the first motion vector differences of the reference blocks, the median of the horizontal component of the first motion vector differences of the reference blocks and the median of the vertical component of the first motion vector differences of the reference blocks, respectively Selecting, or existing motion vector prediction methods can be used.
- block 206 may be implemented as follows.
- the minimum value of the first motion vector differences obtained at block 205 may be adopted as a predictor of the first motion vector of the current block. If there is no first motion vector difference of a particular reference block in the second reference block set, the first motion vector difference of that reference block may not be considered in the calculation process. If there is no first motion vector difference of all the reference blocks in the second reference block set, the predictor of the first motion vector difference of the current block may be set to a zero vector.
- the distance between the frame in which the reference block in the second reference block set is located and the reference frame of the frame in which the reference block in the second reference block set is located is a reference of the frame in which the current block is located and the frame in which the current block is located. It may be different from the distance between the frames.
- the first predictor of the frame and the reference block in the second reference block set First movement of the reference block if not scaled or fully scaled according to the distance between the reference frame of the frame in which the reference block is located in the reference block set and the distance between the frame in which the current block is located and the reference frame of the frame in which the current block is located.
- the vector difference is the distance between the frame in which the reference block in the second reference block set is located and the reference frame in the frame in which the reference block in the second reference block set is located before the predictor of the first motion vector difference of the current block is calculated at block 206.
- reference to the frame where the current block is located and the frame where the current block is located It can be scaled according to the distance between frames.
- a second predictor of the motion vector of the current block is derived according to the first predictor of the motion vector of the current block and the predictor of the first motion vector difference of the current block.
- the method of determining the second predictor of the motion vector of the current block includes performing a sum or weighted sum operation on the first predictor of the motion vector of the current block and the predictor of the first motion vector difference of the current block. .
- the distance between the frame in which the reference block in the second reference block set is located and the reference frame of the frame in which the reference block in the second reference block set is located is a reference of the frame in which the current block is located and the frame in which the current block is located. It may be different from the distance between the frames. Accordingly, in order to improve encoding or decoding performance, if the motion vector of the reference block in the first reference block set and the first predictor of the motion vector of the current block are located, the frame in which the reference block in the first reference block set is located and the first reference may be used.
- the first motion vector difference of the reference block if it is not scaled or fully scaled according to the distance between the reference frame of the frame in which the reference block is located in the block set and the reference frame of the frame in which the current block is located and the frame in which the current block is located. Is the distance between the frame in which the reference block in the first reference block set is located and the reference frame in the frame in which the reference block in the first reference block set is located before the second predictor of the motion vector of the current block is calculated in block 207. True of the frame in which it is located and the frame in which the current block is located It may be scaled based on the distance between the frames.
- the first predictor of the motion vector of the reference block in the second reference block set and the motion vector of the reference block in the second reference block set or the first motion vector of the reference block in the second reference block set is the second of the current block
- the distance between the frame in which the reference block in the first reference block set of the block in the reference block set is located and the reference frame in the frame in which the reference block in the first reference block set of the block in the second reference block set of the current block is located, the second reference block Not scaled or fully scaled according to the distance between the frame in which the reference block in the set is located and the reference frame of the frame in which the reference block in the second reference block is located and the distance between the frame in which the current block is located and the reference frame of the frame in which the current block is located If not, the predictor of the first motion vector difference of the current block In block 207, the distance between the frame in which the reference block in the second reference block set is located and the reference frame in the frame in which the reference block in the second reference block set is located and the frame in which
- the second predictor of the motion vector of the current block is a predictor of the motion vector of the current block, which can be used for encoding or decoding the current block, the current block. It can be adopted as a candidate of the predictor of the motion vector of the motion vector of the current block or sub-block, or a candidate of the motion vector of the current block.
- the second embodiment is described as follows.
- FIG. 4 is a flowchart of a second embodiment according to the present invention.
- 3 is a schematic diagram illustrating reference blocks of a first category of a current block according to an embodiment of the present invention.
- 5 is a schematic diagram illustrating reference blocks of a second category of a current block according to an embodiment of the present invention.
- a method of obtaining motion information is implemented as follows.
- a first reference block set of the current block is determined.
- block 401 may be implemented as follows.
- the first reference block set only includes reference blocks of the first category.
- the reference blocks of the first category of the current block include AL1, AL2, AU1, AU2, AU3, AC1 and AC2.
- the reference blocks in the first reference block set of the current block are the adjacent block AU1 located above the upper left corner of the current block, the adjacent block AL2 located to the left of the upper left corner of the current block, and the upper left corner of the upper left corner of the current block.
- the adjacent block AC2 located at and the adjacent block AU3 located at the upper right of the upper right corner of the current block may be configured.
- the existing method of returning a motion vector can be applied.
- the method of returning a motion vector may be implemented as follows. If block AU3 has a motion vector, the motion vector of block AU3 is returned and decoded into mv3, but the motion vector of block AC2 is not returned. If block AU3 has no motion vector and instead block AC2 has a motion vector, the motion vector of block AC2 is returned and decoded to mv3. If neither block AU3 nor AC2 has a motion vector, a zero vector is returned and decoded to mv3. If the block AU1 has a motion vector, the motion vector of the block AU1 is returned and decoded into mv1.
- block AU2 has a motion vector
- the motion vector of block AU2 is returned and decoded into mv2. If the block AU1 does not have a motion vector, a zero vector is returned and decoded into mv1. If block AU2 does not have a motion vector, a zero vector is returned and decoded into mv2. If a particular block does not exist, the block is considered to have no motion vector.
- the process method described herein may be applied to other embodiments of the present invention.
- the first predictor of the motion vector of the current block is calculated according to the motion vector of each reference block in the first reference block set obtained in block 402.
- the first predictor of the motion vector of the current block may be calculated according to the same method as used in the existing encoding or decoding method for predicting or deriving the motion vector of the current block.
- Block 403 may be implemented as follows.
- the horizontal and vertical components of mv1 are decoded into mv1_x and mv1_y, respectively, and the horizontal and vertical components of mv2 are decoded into mv2_x and mv2_y, respectively, and the horizontal and vertical components of mv3 are decoded into mv3_x and mv3_y, respectively. If mv1_x and mv2_x have the same sign, the mean of mv1_x and mv2_x is taken as the horizontal component of the first predictor of the motion vector of the current block.
- the mean of mv1_x and mv3_x is taken as the horizontal component of the first predictor of the motion vector of the current block. If mv1_x and mv2_x have different symbols and mv2_x and mv3_x have the same symbol, the mean of mv2_x and mv3_x is taken as the horizontal component of the first predictor of the motion vector of the current block.
- the mean of mv1_y and mv2_y is taken as the vertical component of the first predictor of the motion vector of the current block. If mv1_y and mv2_y have different symbols and mv1_y and mv3_y have the same symbol, the mean of mv1_y and mv3_y is taken as the vertical component of the first predictor of the motion vector of the current block.
- the mean of mv2_y and mv3_y is taken as the vertical component of the first predictor of the motion vector of the current block. If the first predictor of the motion vector of the specific reference block does not exist, the motion vector of the reference block is set to a zero vector.
- a difference of the first motion vector of the current block may be calculated.
- the first motion vector difference of the reference block can be obtained immediately.
- the motion vector is scaled at block 403.
- a second reference block set of the current block is determined.
- Block 404 may be implemented as follows.
- the second reference block set includes only reference blocks of the first category and reference blocks of the second category.
- the reference blocks of the first category of the current block include AL1, AL2, AU1, AU2, AU3, AC1 and AC2.
- the reference blocks of the first category in the second reference block set of the current block are the adjacent block AU1 located above the upper left corner of the current block, the adjacent block AL2 located to the left of the upper left corner of the current block, and the upper left corner of the current block. It may consist of an adjacent block AC2 located at the upper left of the vertex.
- the reference blocks of the second category of the current block include B1, B2, B3, B4, B5, B6 and B7.
- B1, B2, B3 and B4 overlap with the current block in space.
- B5, B6 and B7 are adjacent to the current block in space.
- the reference blocks of the second category in the second reference block set of the current block are B1 including a point located in the same space as the upper left corner of the current block, a point located in the same space as the lower right corner of the current block. It may be configured to include B4.
- a first motion vector difference is obtained between the motion vector of each reference block in the second reference block set and the first predictor of the motion vector of the reference block in the second reference block set.
- the first motion vector difference between the motion vector of the current block and the first predictor of the motion vector of the current block is calculated and stored.
- the reference blocks in the second reference block set are all coded blocks. And the first motion vector difference of each reference block in the second reference block set is generated and stored. Therefore, the first motion vector difference of each reference block in the second reference block set can be directly obtained.
- the predictor of the first motion vector difference of the current block at block 406 is calculated according to the first motion vector difference of each reference block in the second reference block set calculated at block 405.
- Block 406 may be implemented as follows.
- the encoder selects one of the existing first motion vector differences and the zero vector as a predictor of the first motion vector difference of the current block.
- the index of the selected value is recorded in the bit stream.
- the decoder selects the correct value from the existing first motion vector differences and zero vector obtained at block 405 according to the index in the bit stream.
- the selected value is adopted as the predictor of the first motion vector difference of the current block.
- the index transmitted in the bit stream is derived from the method of transmitting the derivation information in the bit stream for calculating the predictor of the first motion vector difference in the description of FIG. 1. Can be adopted as information.
- the first motion vector difference may be scaled.
- the second predictor of the motion vector of the current block is derived according to the predictor of the first motion vector difference of the current block and the first predictor of the motion vector of the current block.
- Block 407 is not described herein, but is the same as block 207 of the first embodiment.
- the second predictor of the motion vector of the current block is a predictor of the motion vector of the current block, which can be used for encoding or decoding the current block, the current block. It can be adopted as a candidate of the predictor of the motion vector of the motion vector of the current block or the candidate of the motion vector of the current block.
- the first predictor of the motion vector of the current block is calculated first.
- the predictor of the first motion vector difference of the current block is calculated.
- the first predictor of the motion vector of the current block and the predictor of the first motion vector difference of the current block may be obtained at the same time.
- the predictor of the first motion vector difference of the current block may be calculated first, and the first predictor of the motion vector of the current block may be calculated.
- a predictor of the first motion vector difference of the current block is first calculated, and a first predictor of the motion vector of the current block is calculated.
- the third embodiment is described as follows.
- FIG. 6 is a flowchart of a third embodiment according to the present invention.
- 5 is a schematic diagram illustrating reference blocks of a second category of a current block according to an embodiment of the present invention. As shown in FIG. 6, a method of obtaining motion information is implemented as follows.
- a second set of reference blocks of the current block is determined.
- Block 601 may be implemented as follows.
- the second reference block set only includes reference blocks of the second category.
- the reference blocks of the second category of the current block include B1, B2, B3, B4, B5, B6, and B7.
- B1, B2, B3 and B4 overlap with the current block in space.
- B5, B6 and B7 are adjacent to the current block in space.
- the reference blocks in the second reference block set of the current block are B1 including a point located in the same space as the center point of the current block, B1 including a point located in the same space as the upper left corner of the current block, and the current block. It may be composed of B4 including a point located in the same space as the lower right corner of the.
- a motion vector of each reference block in the second reference block set and a first predictor of the motion vector of the reference block in the second reference block set are obtained.
- the first motion vector difference between the motion vector of the reference block and the first predictor of the motion vector of the reference block is calculated.
- Block 602 is similar to block 205 described in the first embodiment.
- the motion vector of the specific reference block or the motion of the reference block is obtained. If there is no first predictor of the vector, an indication is returned indicating that there is no first motion vector difference of the reference block.
- the scaling operation may be performed at block 602.
- a predictor of the first motion vector difference of the current block is obtained according to the first motion vector difference of each reference block in the second reference block set obtained in block 602.
- the first motion vector difference of the current block may be predicted according to an existing method of predicting motion vectors.
- Block 603 may be implemented as follows.
- the median of the horizontal components of the first motion vector differences of the reference blocks obtained at block 602 and the median of the vertical components of the first motion vector differences of the reference blocks obtained at block 602 are each the first value of the current block. It is adopted as the horizontal component of the predictor of the motion vector difference and the vertical component of the predictor of the first motion vector difference of the current block. If the first motion vector difference of a particular reference block does not exist or is equal to 0, a zero vector is adopted as a predictor of the first motion vector difference of the current block.
- Block 603 may be implemented as follows.
- the encoder selects one of the existing first motion vector differences obtained at block 602 as the predictor of the first motion vector difference of the current block.
- the index of the selected value is recorded in the bit stream.
- the decoder selects the correct value from the existing first motion vector differences obtained at block 602 according to the index in the bit stream.
- the selected value is adopted as the predictor of the first motion vector difference of the current block.
- the index transmitted in the bit stream is derived from the method of transmitting the derivation information in the bit stream for calculating the predictor of the first motion vector difference in the description of FIG. 1. Can be adopted as information.
- a scaling operation may be performed.
- a first reference block set of the current block is determined.
- Block 604 may be implemented as follows.
- the first reference block set only includes reference blocks of the second category.
- the reference blocks of the second category of the current block include B1, B2, B3, B4, B5, B6 and B7.
- B1, B2, B3 and B4 overlap with the current block in space.
- B5, B6 and B7 are adjacent to the current block in space.
- the reference blocks in the second reference block set of the current block are B1 including a point located in the same space as the upper left corner of the current block, and B4 including a point located in the same space as the lower right corner of the current block. It may be configured as.
- the first predictor of the motion vector of the current block at block 606 is calculated according to the motion vector of each reference block in the first reference block set obtained at block 605.
- the first predictor of the motion vector of the current block may be calculated according to the same method as used in the existing encoding or decoding method for predicting or deriving the motion vector of the current block.
- Block 606 may be implemented as follows.
- the encoder selects one of the existing motion vectors as the predictor of the first motion vector difference of the current block.
- the index of the selected value is recorded in the bit stream.
- the decoder selects the correct value from the existing motion vectors obtained at block 605 according to the index in the bit stream.
- the selected value is adopted as the first predictor of the motion vector of the current block. If there is no motion vector of the specific reference block in the first reference block set, the non-existent motion vector of the reference block is set to zero vector.
- the index transmitted in the bit stream is the derivation information of the first predictor of the motion vector of the current block, which is mentioned in the method of transmitting the derivation information in the bit stream for calculating the first predictor of the motion vector in the description of FIG. Can be adopted.
- the scaling operation may be performed at block 606.
- the second predictor of the motion vector of the current block is derived according to the predictor of the first motion vector difference of the current block and the first predictor of the motion vector of the current block.
- Block 607 is identical to block 207 described in the first embodiment.
- the second predictor of the motion vector of the current block is a predictor of the motion vector of the current block, which can be used for encoding or decoding the current block, the current block. It can be adopted as a candidate of the predictor of the motion vector of the motion vector of the current block or the candidate of the motion vector of the current block.
- the fourth embodiment is described as follows.
- FIG. 7 is a flowchart of a fourth embodiment according to the present invention.
- 3 is a schematic diagram illustrating reference blocks of a first category of a current block according to an embodiment of the present invention.
- 5 is a schematic diagram illustrating reference blocks of a second category of a current block according to an embodiment of the present invention.
- a method of obtaining motion information is implemented as follows.
- a second set of reference blocks of the current block is determined.
- Block 701 may be implemented as follows.
- the second reference block set only includes reference blocks of the second category.
- the reference blocks of the second category of the current block include B1, B2, B3, B4, B5, B6, and B7.
- B1, B2, B3 and B4 overlap with the current block in space.
- B5, B6 and B7 are adjacent to the current block in space.
- the reference blocks in the second reference block set of the current block include the upper block B1 including a point located in the same space as the upper left corner of the current block, and a point located in the same space as the upper left corner of the current block.
- the upper left block B1 may include a left block B1 and a point located in the same space as the upper left corner of the current block.
- a first motion vector difference is obtained between the motion vector of each reference block in the second reference block set and the first predictor of the motion vector of the reference block in the second reference block set.
- Block 702 is similar to block 405 described in the second embodiment.
- Block 702 When the first motion vector difference of each reference block in the second reference block set is obtained at block 702, if there is no first motion vector difference of the particular reference block in the second reference block set, An indication is returned indicating that one motion vector difference does not exist.
- the scaling operation may be performed at block 702.
- the predictor of the first motion vector difference of the current block is obtained according to the first motion vector difference of each reference block in the second reference block set obtained in block 702.
- the predictor of the first motion vector difference of the current block may be calculated according to an existing method of predicting the motion vector.
- Block 703 may be implemented as follows.
- the first motion vector difference returned for the upper block is mvd_U
- the first motion vector difference returned for the left block is mvd_L
- the first motion vector difference returned for the upper left block is mvd_LU.
- , d13
- , and d23
- mvd_LU is taken as a predictor of the first motion vector difference of the current block. If d13 is the smallest value among d12, d13, and d23, mvd_U is taken as a predictor of the first motion vector difference of the current block. If d23 is the smallest value among d12, d13, and d23, mvd_L is taken as a predictor of the first motion vector difference of the current block. If the first motion vector difference of the specific reference block does not exist or is equal to 0, the predictor of the first motion vector difference of the current block is set to a zero vector.
- a scaling operation may be performed.
- the first reference block set of the current block is determined.
- Block 704 may be implemented as follows.
- the first reference block set includes only reference blocks of the first category and reference blocks of the second category.
- the reference blocks of the first category of the current block include AL1, AL2, AU1, AU2, AU3, AC1 and AC2.
- the reference blocks of the second category of the current block include B1, B2, B3, B4, B5, B6 and B7.
- B1, B2, B3 and B4 overlap with the current block in space.
- B5, B6 and B7 are adjacent to the current block in space.
- the reference blocks in the first reference block set of the current block are adjacent block AC2 located at the upper left corner of the upper left corner of the current block, adjacent block AU3 located at the upper right corner of the upper right corner of the current block, and left of the current block. It may be composed of a neighboring block AC1 located at the bottom left of the bottom vertex and a B1 including a point located in the same space as the center point of the current block.
- the first predictor of the motion vector of the current block at block 706 is calculated according to the motion vector of each reference block in the first reference block set obtained at block 705.
- the first predictor of the motion vector of the current block may be calculated according to the same method as used in the existing encoding or decoding method for predicting or deriving the motion vector of the current block.
- block 706 may be implemented as follows.
- the average value of the motion vector mv1 of AC1, the motion vector mv2 of AC2, and the motion vector mv3 of AC3 is calculated and decoded by mva. If any one of mv1, mv2 and mv3 does not exist, then the nonexistent one is set to a zero vector.
- the motion vector of B1 is decoded with mvb. The smaller of mva and mvb is adopted as the first predictor of the motion vector of the current block. If mvb does not exist, mva is taken as the first predictor of the motion vector of the current block.
- the first motion vector difference of the current block is calculated.
- the first motion vector difference of the reference block can be obtained directly.
- a scaling operation may be performed at block 706.
- the second predictor of the motion vector of the current block is derived according to the first predictor of the motion vector of the current block and the predictor of the first motion vector difference of the current block.
- Block 707 is similar to block 207 of the first embodiment.
- the second predictor of the motion vector of the current block is a predictor of the motion vector of the current block, which can be used for encoding or decoding the current block, the current block. It can be adopted as a candidate of the predictor of the motion vector of the motion vector of the current block or the candidate of the motion vector of the current block.
- the fifth embodiment is described as follows.
- FIG. 8 is a flowchart of a fifth embodiment of the present invention.
- 9 is a schematic diagram depicting reference blocks of a first category of the current block according to an embodiment of the present invention.
- 10 is a schematic diagram depicting reference blocks of a second category of the current block according to an embodiment of the present invention.
- a method of obtaining motion information is implemented as follows.
- the current block E referenced in FIG. 9 is divided into four sub-blocks E1, E2, E3, and E4.
- the current block is divided into four sub blocks.
- motion information of each subblock can be calculated according to the method described in the present invention.
- motion information of four sub-blocks is obtained according to the following blocks.
- a first reference block set of each current subblock is determined.
- the reference blocks in the first reference block set may be identical, partially identical, or completely different.
- Block 802 may be implemented as follows.
- the sub blocks have different first reference block sets. However, all first reference block sets contain only reference blocks of the second category.
- reference blocks of the second category of E1, E2, E3, and E4 include B1, B2, B3, B4, and B5.
- the reference blocks in the first reference block set of the current subblock may be set to a block including a point having a spatial position equal to the center of the current subblock.
- the first reference block set of E1 is B1
- the first reference block set of E2 is B3
- the first reference block set of E3 is B1
- the first reference block set of E4 is B4.
- the reference block A is the sub block.
- the reference blocks in the first reference block set may be blocks or subblocks. If a single piece of motion information corresponds to a block, the corresponding reference block is a block. If a single piece of motion information corresponds to a sub block, the corresponding reference block is a sub block.
- block A is a coded block proximate to the current subblock.
- block A is divided into sub-blocks A1 and A2, and motion information of sub-block A1 and motion information of sub-block A2 are respectively calculated.
- A1 is close to the current subblock or a block of the current subblock.
- A2 is not close to the current subblock or the block of the current subblock. (Because block A does not correspond to a single piece of motion information, but corresponds to motion information of sub blocks A1 and A2.) Because sub block A2 is not in close proximity to the current sub block and the block of the current sub block, the sub Block A2 is not included in the first reference block set of the current sub block.
- the method of obtaining the motion vector of each reference block in the first reference block set of each current subblock may be the same as the existing method.
- the motion vector of each reference block in the first reference block set of each current sub-block is obtained at block 803, if the motion vector of the specific reference block in the first reference block set does not exist, the motion vector of the reference block An indication is returned indicating that does not exist.
- a first predictor of the motion vector of each current subblock is calculated according to the motion vector of each reference block in the first reference block set of each current subblock obtained at block 803.
- the method of determining the first predictor of the motion vector of each current subblock according to the motion vector of the reference block may be the same as the existing method of predicting or deriving the motion vector.
- the methods for determining the first predictors of the motion vectors of the current subblocks may be the same or different.
- Block 804 may be implemented as follows.
- the motion vector of the reference block in the first reference block set of each current subblock may be directly adopted as the first predictor of the motion vector of each current subblock. If the reference block in the first reference block set of the specific current subblock does not have a motion vector, the first predictor of the motion vector of the current subblock is set to a zero vector, or the second predictor of the motion vector of the current subblock is blocks.
- the predictor of the motion vector of the current subblock, the candidate of the predictor of the motion vector of the current subblock, the motion vector of the current subblock, or the candidate of the motion vector of the current subblock can be adopted as.
- the motion vector of the reference block in the first reference block set of each current subblock may be scaled.
- a second reference block set of each current subblock is determined.
- Block 805 may be implemented as follows.
- E1, E2, E3 and E4 have the same second reference block sets.
- each second reference block set includes only reference blocks of the first category.
- reference blocks of the first category of E1, E2, E3 and E4 include AL1, AC1, AU1, AU2 and AU3.
- the reference blocks of the first set of reference blocks of E1, E2, E3, and E4 are located at the upper left corner of the blocks AU1 and E1 above the upper left corner of the blocks AL1, E1, which are located to the left of the upper left corner of E1. It may consist of the block AC1 in the upper left corner.
- block AL1 located to the left of the upper left corner of E1 has no motion vector
- block AL1 located to the left of the upper left corner of E1 is replaced with block AL1 located to the left of the upper left corner of E3.
- block AU1 reference motion vector located above the upper left corner of E1 is replaced with block AU2 located above the upper left corner of E2.
- the first predictor of the motion vector of each reference block in the second reference block set and the motion vector of the reference block in the second reference block set is obtained for each current subblock.
- the first motion vector difference between the motion vector of the reference block and the first predictor of the motion vector of the reference block is calculated.
- Block 806 is similar to block 205 described in the first embodiment, but differs from block 205 in that each current subblock is processed separately. However, in this embodiment, the second reference block sets of all current subblocks are the same, so the process only needs to be performed once.
- the first predictor of the motion vector of each reference block in the second reference block set of each current subblock and the motion vector of each reference block in the second reference block set of each current subblock If there is no motion vector of the specific reference block or the first predictor of the motion vector of the reference block, an indication is returned indicating that there is no first motion vector difference of the reference block.
- the motion vector of the reference block in the second reference block set and the first predictor of the motion vector of the reference block in the second reference block set may be scaled for each current subblock of this block.
- the predictor of the first motion vector difference of each current subblock is calculated according to the first motion vector difference of each reference block in the second reference block set of the current subblock calculated at block 806.
- Predictors of the first motion vector differences of the current subblocks may be calculated through the same method or other methods. The calculation method may be determined according to the actual requirements. For example, one of the methods mentioned in the above embodiment may be applied.
- block 807 may be implemented as follows. The average of existing first motion vector differences of the reference blocks in the second reference block set of the current subblock may be adopted as a predictor of the first motion vector difference of the current subblock. If all existing first motion vector differences of the reference blocks in the second reference block set of the specific current subblock do not exist, the predictor of the first motion vector difference of the current subblock may be set to zero vector.
- Block 807 may also be implemented as follows.
- the encoder selects one of the existing first motion vector differences of the reference blocks in the second reference block set of the current subblock calculated at block 806, and predicts the selected value of the first motion vector difference of the current subblock.
- the index of the selected value is recorded in the bit stream.
- the decoder selects the correct value according to the index in the bit stream from the existing first motion vector differences of the reference blocks in the second reference block set of the current sub-block calculated at block 806, and selects the selected value of the current sub-block.
- a predictor of the first motion vector difference is determined.
- the transmitted index in the bit stream is derived from the predictor of the first motion vector difference of the current sub-block mentioned in the method of transmitting the derived information in the bit stream to calculate the predictor of the first motion vector difference in the description of FIG. Can be adopted as.
- the subblocks can share the predictor of the first motion vector difference. In this case, the calculation process of the predictor of the first motion vector difference must be performed only once for the entire current block E, and only one index of the first motion vector difference is transmitted into the bit stream.
- the first motion vector difference of the reference blocks in the second reference block set may be scaled with respect to the current subblock in each of these blocks.
- a second predictor of the motion vector of the current subblock is derived according to the first predictor of the motion vector of the current subblock and the predictor of the first motion vector difference of the current subblock.
- Block 808 is similar to block 207 described in the first embodiment, and differs from block 207 in that each current subblock is processed separately.
- the second predictor of the motion vector of the current subblock is the predictor of the motion vector of the current subblock used in the encoding or decoding process of the current subblock.
- the sixth embodiment is as follows.
- FIG. 11 is a flowchart of a sixth embodiment of the present invention.
- 12 is a schematic diagram illustrating reference blocks of a first category of a current block according to an embodiment of the present invention.
- 13 is a schematic diagram depicting reference blocks of a second category of the current block according to an embodiment of the present invention.
- a method for obtaining motion information is implemented as follows.
- the current block E is divided into E1 and E2.
- the current block is divided into E1 and E2.
- the following blocks 1102-1108 are performed for E1 and E2, respectively.
- the second reference block set of the current subblock is determined.
- Block 1102 may be implemented as follows.
- E1 and E2 have different second reference block sets. However, each second reference block set only includes reference blocks of the second category.
- the reference blocks of the second category of E1 and E2 include B1, B2, B3, B4, and B5.
- B1, B2, and B3 overlap block E in space.
- B4 and B5 are close to block E in space.
- the reference blocks in the first reference block set of the current subblock may be set to a block including a point having a spatial position equal to the center of the current subblock.
- B1 may be a reference block in the second reference block set of the current subblock.
- B3 may be a reference block in the second reference block set of the current subblock.
- a motion vector of each reference block in the second reference block set of the current subblock and a first predictor of the motion vector of the reference block in the second reference block set of the current subblock are obtained.
- the first motion vector difference between the motion vector of the reference block and the first predictor of the motion vector of the reference block is calculated.
- the motion vector and the first motion vector difference can be scaled to fit this block.
- the predictor of the first motion vector difference of the current subblock is calculated according to the first motion vector difference of each reference block in the second reference block set of the current subblock obtained at block 1103.
- Block 1104 may be implemented as follows.
- the first motion vector difference of the reference block in the second reference block set of the current subblock obtained in block 1103 may be directly adopted as a predictor of the first motion vector difference of the current subblock. If there is no first motion vector difference of a particular subblock in the second reference block set, the predictor of the first motion vector difference of the current subblock is set to a zero vector.
- the first motion vector difference in this block can be scaled.
- a first motion vector difference of the current subblock is determined.
- Block 1105 may be implemented as follows.
- E1 and E2 have different first reference block sets. However, each first reference block set only includes reference blocks of the first category.
- the reference blocks of the first category of E1 and E2 include AL1, AL2, AC1, AU1, AU2 and AU3.
- the reference blocks of the first reference block set of the current subblock may be composed of a block located to the left of the upper left corner of the current subblock and a block above the upper left corner of the current subblock. For example, if the current subblock is E1, AL2 and AU1 are the reference blocks in the first reference block set of the current subblock. If the current sub block is E2, E1 and AU2 are reference blocks in the first reference block set of the current sub block.
- the motion vector of each reference block in the first reference block set of the current sub-block is obtained at block 1106, if the motion vector of the specific reference blocks in the first reference block set does not exist, then the motion vector of the reference block is An indication is returned indicating that it does not exist.
- the first predictor of the motion vector of the current sub-block is obtained according to the motion vector of each reference block in the first reference block set of the current sub-block in block 1106.
- Block 1107 may be implemented as follows.
- the minimum value of the motion vectors of the reference blocks in the first reference block set of the current subblock may be adopted as the first predictor of the motion vector of the current subblock. If the motion vectors of the reference blocks in the first reference block set of the current subblock are not all present, the first predictor of the motion vector of the current subblock is set to a zero vector.
- the motion vectors of the reference blocks in the first reference block set in this block may be scaled.
- a second predictor of the motion vector of the current subblock is obtained according to the first predictor of the motion vector of the current subblock and the predictor of the first motion vector difference of the current subblock.
- Block 1108 is similar to block 207 described in the first embodiment. The difference between the block 207 and the process performed in the current sub-block is different.
- the second predictor of the motion vector of the current subblock is obtained from the motion vector of the current subblock that can be used in the encoding or decoding process of the current subblock.
- a predictor, a candidate of a predictor of the first motion vector difference of the current subblock, a motion vector of the current subblock, or a candidate of the motion vector of the current subblock may be adopted.
- This embodiment differs from the fifth embodiment in that the subblocks of the current block are processed sequentially, but the fifth embodiment is processed at the same time.
- the seventh embodiment is described as follows.
- FIG. 14 is a schematic diagram illustrating an apparatus for obtaining motion information according to an embodiment of the present invention.
- Figure 15 is a schematic diagram illustrating an input of the device according to an embodiment of the present invention.
- 16 is a schematic diagram illustrating a prediction unit of the apparatus according to an embodiment of the present invention.
- an apparatus for obtaining motion information in an encoder / decoder or a video encoding or decoding system is provided according to an embodiment.
- the apparatus includes an input unit 141, a predictor 142, and a calculator 143.
- the input unit 141 inputs motion information of reference blocks in the first reference block set and the second reference block set of the current block or subblock.
- the input unit 141 may be implemented as follows. As shown in FIG. 15, the input unit 141 includes a first reference block set information input unit and a second reference block set information input unit.
- the first reference block set information input unit inputs a motion vector of a reference block in the first reference block set of the current block or subblock.
- the second reference block set information input section includes a first reference between the motion vector of the reference block in the second reference block set of the current block or subblock and the first predictor of the motion vector of the reference block in the second reference block set of the current block or subblock. 1 Enter the motion vector difference.
- the second reference block set information input unit is configured to obtain the first prediction information of the motion vector of the reference block in the second reference block set of the current block or subblock and the motion vector of the reference block in the second reference block set of the current block or subblock. Input, and calculate a first motion vector difference of the reference block in the second reference block set of the current block or sub-block.
- the predictor 142 calculates a first predictor of the motion vector of the current block or subblock and a predictor of the difference of the first motion vector of the current block or subblock.
- the prediction unit 142 may be implemented as follows. As shown in FIG. 16, the predictor 142 calculates a first predictor that calculates a first predictor of a motion vector of a current block or a subblock, and a predictor that calculates a predictor of a difference between a first motion vector of a current block or a subblock. It includes 2 predictors.
- the calculation unit 143 calculates a second predictor of the motion vector of the current block or the subblock according to the first predictor of the motion vector of the current block or the subblock and the predictor of the difference of the first motion vector of the current block or the subblock. do. After the second predictor of the motion vector of the current block or subblock is obtained, the calculator 143 may use the second predictor of the motion vector of the current block or subblock to encode or decode the current block or subblock. It may be adopted as a predictor of the motion vector of the current block or subblock, a candidate of the predictor of the motion vector of the current block or subblock, a motion vector of the current block or subblock, or a candidate of the motion vector of the current block or subblock.
- the motion vector of the current block or subblock is predicted according to the reference blocks of the first reference block set to obtain a first predictor of the motion vector of the current block or subblock.
- the first motion vector difference of the current block or subblock is predicted according to the reference blocks of the second reference block set to obtain a predictor of the first motion vector difference of the current block or subblock.
- the second predictor of the motion vector of the current block or subblock is obtained according to the first predictor of the motion vector of the current block or subblock and the predictor of the difference of the first motion vector of the current block or subblock.
- the second predictor is a predictor of a motion vector of the current block or subblock, a candidate of a predictor of the motion vector of the current block or subblock, a motion vector of the current block or subblock, which can be used for encoding or decoding the current block, or It may be adopted as a candidate of the motion vector of the current block or subblock.
- the first predictor of the motion vector is the final predictor of the motion vector of the current block or subblock, the candidate of the predictor of the motion vector of the current block or subblock, the motion vector or the current block of the current block or subblock. Or as a candidate of the motion vector of the subblock. Based on the existing method of predicting motion vectors, auxiliary prediction of motion vectors has been introduced.
- the accuracy of the predicted or derived motion vectors is improved through the predictors of the motion vector differences.
- the transmitted motion vector difference is reduced. Therefore, the compression efficiency of the motion vector is improved.
- a motion vector optimized for the second encoding or decoding method mentioned in the background art is derived, video encoding performance is improved.
- the first reference block set includes only reference blocks of the first category and the second reference block set only includes reference blocks of the second category or the first reference block set only includes reference blocks of the second category
- the correlation of the motion vector in both the spatial domain and the temporal domain can be eliminated without increasing the complexity of the two predictions. That is, a larger coding performance gain can be obtained at a relatively lower cost.
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
Description
Claims (12)
- 비디오 부호화 또는 복호화에 있어서 움직임 정보를 획득하기 위한 방법은 비디오 이미지의 현재 블록 또는 서브 블록의 움직임 정보를 결정하는 방법을 포함하며, 상기 움직임 정보를 결정하는 방법은상기 현재 블록 또는 서브 블록의 제 1 참조 블록 세트 안의 각 참조 블록의 움직임 벡터에 따라 상기 현재 블록 또는 서브 블록의 움직임 벡터의 제 1 예측자를 계산하는 단계;상기 현재 블록 또는 서브 블록의 제 2 참조 블록 세트 안의 각 참조 블록의 움직임 벡터와 상기 현재 블록 또는 서브 블록의 상기 제 2 참조 블록 세트 안의 상기 참조 블록의 상기 움직임 벡터의 제 1 예측자 간의 제 1 움직임 벡터 차이를 결정하고, 상기 현재 블록 또는 서브 블록의 상기 제 1움직임 벡터 차이의 예측자를 획득하기 위하여 상기 제 2 참조 블록 세트 안의 각 참조 블록의 상기 제 1 움직임 벡터 차이에 따라 상기 현재 블록 또는 서브 블록의 상기 움직임 벡터와 상기 현재 블록 또는 서브 블록의 상기 움직임 벡터의 상기 제 1예측자 간의 제 1 움직임 벡터 차이를 예측하는 단계; 및상기 현재 블록 또는 서브 블록의 상기 제 1 움직임 벡터 차이의 상기 예측자와 상기 현재 블록 또는 서브 블록의 상기 움직임 벡터의 상기 제 1 예측자에 따라 상기 현재 블록 또는 서브 블록의 상기 움직임 벡터의 제 2 예측자를 결정하는 단계를 포함하고,상기 제 1 참조 블록 세트는 최초로 상기 현재 블록 또는 서브 블록의 상기 움직임 벡터를 예측하거나 유도하기 위한 참조 블록들을 포함하고; 및상기 제 2 참조 블록 세트는 상기 현재 블록 또는 서브 블록의 상기 제 1 움직임 벡터 차이를 예측하기 위한 참조 블록들을 포함하는 것을 특징으로 하는 움직임 정보를 획득하기 위한 방법.
- 제 1 항에 있어서,상기 현재 블록의 상기 제 1 참조 블록 세트는 현재 프레임 안에서 상기 현재 블록에 인접한 부호화된 블록 또는 복호화된 블록을 포함하거나, 상기 현재 프레임의 참조 프레임 안에서 공간적으로 상기 현재 블록에 인접하거나 겹치는 부호화된 블록 또는 복호화된 블록을 포함하고;상기 현재 블록의 상기 제 2 참조 블록 세트는 상기 현재 블록에 인접한 상기 현재 프레임에서 부호화된 블록 또는 복호화된 블록을 포함하거나, 상기 현재 프레임의 상기 참조 프레임 안에서 공간적으로 상기 현재 블록에 인접하거나 겹치는 부호화된 블록 또는 복호화된 블록을 포함하고;상기 현재 서브 블록의 상기 제 1 참조 블록 세트는 상기 현재 서브 블록 또는 상기 현재 서브 블록의 상기 블록에 인접한 상기 현재 프레임에서 부호화된 블록 또는 복호화된 블록을 포함하거나, 공간적으로 상기 현재 서브 블록 또는 상기 현재 서브 블록의 상기 블록에 인접하거나 겹치는 상기 현재 프레임의 상기 참조 프레임에서 부호화된 블록 또는 복호화된 블록을 포함하고; 및상기 현재 서브 블록의 상기 제 2 참조 블록 세트는 상기 현재 서브 블록 또는 상기 현재 서브 블록의 상기 블록에 인접한 상기 현재 프레임에서 부호화된 블록 또는 복호화된 블록을 포함하거나, 공간적으로 상기 현재 서브 블록 또는 상기 현재 서브 블록의 상기 블록에 인접하거나 겹치는 상기 현재 프레임의 상기 참조 프레임에서 부호화된 블록 또는 복호화된 블록을 포함하는 움직임 정보를 획득하기 위한 방법.
- 제 1 항에 있어서,상기 현재 블록 또는 서브 블록의 상기 움직임 벡터의 상기 제 1 예측자를 계산하는 단계에 있어서,부호기에 의하여, 상기 현재 블록 또는 서브 블록의 상기 제 1 참조 블록 세트 안의 각 참조 블록의 상기 모션 벡터에 따라 상기 현재 블록 또는 서브 블록의 상기 움직임 벡터의 상기 제 1 예측자를 유도하고, 비트 스트림(bit-stream)에 상기 현재 블록 또는 서브 블록의 상기 움직임 벡터의 상기 제 1 예측자의 유도 정보를 기록하는 단계; 및복호기에 의하여, 상기 현재 블록 또는 서브 블록의 상기 제 1 참조 블록 세트 안의 각 참조 블록의 상기 모션 벡터에 따라 상기 현재 블록 또는 서브 블록의 상기 움직임 벡터의 상기 제 1 예측자와 상기 비트 스트림 안의 상기 현재 블록 또는 서브 블록의 상기 움직임 벡터의 상기 제 1 예측자의 상기 유도 정보를 유도하는 단계를 포함하는 움직임 정보를 획득하기 위한 방법.
- 제 1 항에 있어서,상기 제 2 참조 블록 세트 안의 각 참조 블록의 상기 움직임 벡터와 상기 현재 블록 또는 서브 블록의 상기 제 2 참조 블록 세트 안의 상기 참조 블록의 상기 움직임 벡터의 상기 제 1 예측자 간의 상기 제 1 움직임 벡터 차이를 결정하는 단계에 있어서,상기 현재 블록 또는 서브 블록의 상기 움직임 정보가 결정된 때, 상기 제 2 참조 블록 세트 안의 각 참조 블록의 상기 움직임 벡터와 상기 제 2 참조 블록 세트 안의 상기 참조 블록의 상기 움직임 벡터의 제 1 예측자를 획득하고, 상기 현재 블록 또는 서브 블록의 상기 제 2 참조 블록 세트의 상기 참조 블록의 상기 제 1 움직임 벡터 차이로 상기 제 2 참조 블록 세트 안의 상기 참조 블록의 상기 움직임 벡터와 상기 제 2 참조 블록 세트 안의 상기 참조 블록의 상기 움직임 벡터의 상기 제 1 예측자 간의 차이를 선택하는 단계상기 현재 블록 또는 서브 블록의 제 2 참조 블록 세트 안의 각 참조 블록에 대하여, 상기 참조 블록의 움직임 정보가 결정된 후, 상기 참조 블록의 다음에 처리되는 블록 또는 서브 블록의 움직임 정보가 결정되기 전에, 상기 참조 블록의 상기 제 1 움직임 벡터 차이로 상기 참조 블록의 움직임 벡터와 상기 참조 블록의 상기 움직임 벡터의 제 1 예측자 간의 차이를 선택하고, 상기 참조 블록의 상기 제 1 움직임 벡터 차이를 저장하는 단계; 및상기 현재 블록 또는 서브 블록의 상기 움직임 정보가 결정되면, 상기 제 2 참조 블록 세트 안의 각 참조 블록의 저장된 제 1 움직임 벡터 차이를 획득하는 단계를 포함하는 움직임 정보를 획득하기 위한 방법.
- 제 4 항에 있어서,상기 현재 블록 또는 서브 블록의 상기 제 2 참조 블록 세트의 참조 블록의 움직임 벡터 또는 상기 현재 블록 또는 서브 블록의 상기 제 2 참조 블록 세트의 상기 참조 블록의 상기 움직임 벡터의 예측자가 존재하지 않을 때, 상기 움직임 정보를 획득하기 위한 방법은,상기 참조 블록의 제 1 움직임 벡터 차이는 상기 현재 블록 또는 서브 블록의 제 1 움직임 벡터 차이의 상기 예측자의 계산에 사용되지 않을 때, 상기 참조 블록의 상기 제 1 움직임 벡터 차이가 존재하지 않는다고 판단하는 단계; 또는상기 참조 블록의 제 1 움직임 벡터 차이는 상기 현재 블록 또는 서브 블록의 제 1 움직임 벡터 차이의 상기 예측자의 계산에 사용될 때, 상기 참조 블록의 상기 제 1 움직임 벡터 차이를 영 벡터로 설정하는 단계를 포함하는 움직임 정보를 획득하기 위한 방법.
- 제 5 항에 있어서,상기 현재 블록 또는 서브 블록의 상기 제 2 참조 블록 세트 안의 상기 참조 블록의 상기 제 1 움직임 벡터 차이가 존재하지 않을 때, 상기 움직임 정보를 획득하기 위한 방법은 상기 현재 블록 또는 서브 블록의 상기 제 1 움직임 벡터 차이의 상기 예측자를 영 벡터로 설정하는 단계를 포함하거나,상기 현재 블록 또는 서브 블록의 상기 제 2 참조 블록 세트 안의 모든 참조 블록들의 제 1 움직임 벡터 차이들이 존재하지 않을 때, 상기 움직임 정보를 획득하기 위한 방법은 상기 현재 블록 또는 서브 블록의 상기 제 1 움직임 벡터 차이의 상기 예측자를 영 벡터로 설정하는 단계를 포함하는 것을 특징으로 하는 움직임 정보를 획득하기 위한 방법.
- 제 1 항에 있어서,상기 현재 블록 또는 서브 블록의 제 1 움직임 벡터 차이를 예측하는 단계는,상기 제 2 참조 블록 세트의 상기 참조 블록들의 제 1 움직임 벡터 차이들의 수평 성분들 중 최소값을 상기 현재 블록 또는 서브 블록의 상기 제 1 움직임 벡터 차이의 상기 예측자의 수평 성분으로 채택하고, 상기 제 2 참조 블록 세트의 상기 참조 블록들의 제 1 움직임 벡터 차이들의 수직 성분들 중 최소값을 상기 현재 블록 또는 서브 블록의 상기 제 1 움직임 벡터 차이의 상기 예측자의 수직 성분으로 채택하는 단계; 또는상기 제 2 참조 블록 세트의 상기 참조 블록들의 제 1 움직임 벡터 차이들의 평균을 상기 현재 블록 또는 서브 블록의 상기 제 1 움직임 벡터 차이의 상기 예측자로 채택하는 단계; 또는상기 제 2 참조 블록 세트의 상기 참조 블록들의 제 1 움직임 벡터 차이들의 수평 성분들 중 중간값을 상기 현재 블록 또는 서브 블록의 상기 제 1 움직임 벡터 차이의 상기 예측자의 수평 성분으로 채택하고, 상기 제 2 참조 블록 세트의 상기 참조 블록들의 제 1 움직임 벡터 차이들의 수직 성분들 중 중간값을 상기 현재 블록 또는 서브 블록의 상기 제 1 움직임 벡터 차이의 상기 예측자의 수직 성분으로 채택하는 단계; 또는상기 제 2 참조 블록 세트의 상기 참조 블록들의 제 1 움직임 벡터 차이들 중 각 두 개 사이의 거리를 계산하고, 상기 거리들에 따라 상기 현재 블록 또는 서브 블록의 제 1 움직임 벡터 차이의 상기 예측자를 유도하는 단계; 또는부호기에 의하여, 상기 제 2 참조 블록 세트의 각 참조 블록의 상기 제 1 움직임 벡터 차이에 따라 상기 현재 블록 또는 서브 블록의 제 1 움직임 벡터 차이의 상기 예측자를 유도하고, 상기 현재 블록 또는 서브 블록의 제 1 움직임 벡터 차이의 상기 예측자의 유도 정보를 비트 스트림에 기록하는 단계; 및 복호기에 의하여, 상기 제 2 참조 블록 세트 안의 각 참조 블록의 상기 제 1 움직임 벡터 차이에 따라 상기 현재 블록 또는 서브 블록의 상기 제 1 움직임 벡터 차이의 상기 예측자, 및 상기 현재 블록 또는 서브 블록의 상기 제 1 움직임 벡터 차이의 상기 예측자의 상기 유도 정보를 유도하는 단계를 포함하는 움직임 정보를 획득하기 위한 방법.
- 제 1 항에 있어서,상기 현재 블록 또는 서브 블록의 상기 움직임 벡터의 상기 제 2 예측자를 결정하는 단계는,상기 현재 블록 또는 서브 블록의 상기 제 1 움직임 벡터 차이의 상기 예측자와 상기 현재 블록 또는 서브 블록의 상기 움직임 벡터의 상기 제 1 예측자의 합 또는 상기 현재 블록 또는 서브 블록의 상기 제 1 움직임 벡터 차이의 상기 예측자와 상기 현재 블록 또는 서브 블록의 상기 움직임 벡터의 상기 제 1 예측자의 가중합을 상기 현재 블록 또는 서브 블록의 상기 움직임 벡터의 상기 제 2 예측자로 채택하는 단계를 포함하는 움직임 정보를 획득하기 위한 방법.
- 제 1 항 내지 제8항에 있어서,상기 현재 블록 또는 서브 블록의 상기 제 1 참조 블록 세트 안의 상기 참조 블록들은 상기 현재 블록 또는 서브 블록의 상기 제 2 참조 블록 세트 안의 상기 참조 블록들과 동일하거나 다른 것을 특징으로 하는 움직임 정보를 획득하기 위한 방법.
- 제 1 항 내지 제8항에 있어서,블록 내에서 다른 서브 블록들에 대하여, 서브 블록의 제 1 참조 블록 세트 안의 참조 블록들은 또 다른 서브 블록의 제 1 참조 블록 세트 안의 참조 블록들과 동일하거나 다르고, 서브 블록의 제 2 참조 블록 세트 안의 참조 블록들은 또 다른 서브 블록의 제 2 참조 블록 세트 안의 참조 블록들과 동일하거나 다른 것을 특징으로 하는 움직임 정보를 획득하기 위한 방법.
- 제 1 항 내지 제8항에 있어서,상기 제 1 참조 블록 세트 또는 상기 제 2 참조 블록 세트 안의 참조 블록의 움직임 정보가 서브 블록 안에서 결정될 때, 상기 참조 블록은 서브 블록인 것을 특징으로 하는 움직임 정보를 획득하기 위한 방법.
- 비디오 부호화 또는 복호화에 관한 움직임 정보를 획득하기 위한 장치에 있어서,현재 블록 또는 서브 블록의 제 1 참조 블록 세트 안의 각 참조 블록의 움직임 벡터를 획득하고, 상기 현재 블록 또는 서브 블록의 제 2 참조 블록 세트 안의 각 참조 블록과 상기 현재 블록 또는 서브 블록의 상기 제 2 참조 블록 세트 안의 상기 참조 블록의 상기 움직임 벡터의 제 1 예측자 간의 제 1 움직임 벡터 차이를 결정하는 입력부;상기 제 1 참조 블록 세트 안의 각 참조 블록의 상기 움직임 벡터에 따라 상기 현재 블록 또는 서브 블록의 상기 움직임 벡터의 상기 제 1 예측자를 계산하고, 상기 현재 블록 또는 서브 블록의 상기 제 1 움직임 벡터 차이의 상기 예측자를 획득하기 위하여 상기 제 2 참조 블록 세트 안의 각 참조 블록의 상기 제 1 움직임 벡터 차이에 따라 상기 현재 블록 또는 서브 블록의 상기 제 1 움직임 벡터 차이를 예측하는 예측부; 및상기 현재 블록 또는 서브 블록의 상기 움직임 벡터의 상기 제 1 예측자에 따라 상기 현재 블록 또는 서브 블록의 상기 움직임 벡터의 제 2 예측자를 결정하는 계산부를 포함하고,상기 제 1 참조 블록 세트는 최초로 상기 현재 블록 또는 서브 블록의 상기 움직임 벡터를 예측하거나 유도하기 위한 참조 블록들을 포함하고; 및상기 제 2 참조 블록 세트는 상기 현재 블록 또는 서브 블록의 상기 제 1 움직임 벡터 차이를 예측하기 위한 참조 블록들을 포함하는 것을 특징으로 하는 움직임 정보를 획득하기 위한 장치.
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
US14/898,291 US10051282B2 (en) | 2013-06-14 | 2014-06-13 | Method for obtaining motion information with motion vector differences |
SG11201510258PA SG11201510258PA (en) | 2013-06-14 | 2014-06-13 | Method for obtaining motion information |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201310236309 | 2013-06-14 | ||
CN201310236309.6 | 2013-06-14 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2014200313A1 true WO2014200313A1 (ko) | 2014-12-18 |
Family
ID=52022517
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/KR2014/005226 WO2014200313A1 (ko) | 2013-06-14 | 2014-06-13 | 움직임 정보를 획득하기 위한 방법 |
Country Status (5)
Country | Link |
---|---|
US (1) | US10051282B2 (ko) |
KR (1) | KR20140146541A (ko) |
CN (2) | CN111200736B (ko) |
SG (1) | SG11201510258PA (ko) |
WO (1) | WO2014200313A1 (ko) |
Families Citing this family (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
EP4221202A1 (en) * | 2015-06-05 | 2023-08-02 | Dolby Laboratories Licensing Corporation | Image encoding and decoding method and image decoding device |
CN116193109A (zh) * | 2017-01-16 | 2023-05-30 | 世宗大学校产学协力团 | 影像编码/解码方法 |
CN107071421B (zh) * | 2017-05-23 | 2019-11-22 | 北京理工大学 | 一种结合视频稳定的视频编码方法 |
CN109005412B (zh) * | 2017-06-06 | 2022-06-07 | 北京三星通信技术研究有限公司 | 运动矢量获取的方法及设备 |
US10856003B2 (en) * | 2017-10-03 | 2020-12-01 | Qualcomm Incorporated | Coding affine prediction motion information for video coding |
CN109996075B (zh) * | 2017-12-29 | 2022-07-12 | 华为技术有限公司 | 一种图像解码方法及解码器 |
WO2019199067A1 (ko) * | 2018-04-13 | 2019-10-17 | 엘지전자 주식회사 | 비디오 처리 시스템에서 인터 예측 방법 및 장치 |
US11051025B2 (en) * | 2018-07-13 | 2021-06-29 | Tencent America LLC | Method and apparatus for video coding |
CN111050177B (zh) * | 2018-10-15 | 2021-07-20 | 腾讯科技(深圳)有限公司 | 视频编码、视频解码方法、装置、计算机设备和存储介质 |
CN112866699B (zh) | 2019-03-11 | 2022-11-01 | 杭州海康威视数字技术股份有限公司 | 一种编解码方法、装置及其设备 |
CN110719489B (zh) * | 2019-09-18 | 2022-02-18 | 浙江大华技术股份有限公司 | 运动矢量修正、预测、编码方法、编码器及存储装置 |
CN113706573B (zh) * | 2020-05-08 | 2024-06-11 | 杭州海康威视数字技术股份有限公司 | 一种运动物体的检测方法、装置及存储介质 |
CN111626178B (zh) * | 2020-05-24 | 2020-12-01 | 中南民族大学 | 一种基于新时空特征流的压缩域视频动作识别方法和系统 |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20060153300A1 (en) * | 2005-01-12 | 2006-07-13 | Nokia Corporation | Method and system for motion vector prediction in scalable video coding |
US20110080954A1 (en) * | 2009-10-01 | 2011-04-07 | Bossen Frank J | Motion vector prediction in video coding |
KR20120026092A (ko) * | 2009-06-26 | 2012-03-16 | 후아웨이 테크놀러지 컴퍼니 리미티드 | 비디오 이미지의 움직임 정보를 취득하는 방법, 장치, 및 기기, 그리고 템플릿 구성 방법 |
US20120128060A1 (en) * | 2010-11-23 | 2012-05-24 | Mediatek Inc. | Method and Apparatus of Spatial Motion Vector Prediction |
US20130003851A1 (en) * | 2011-07-01 | 2013-01-03 | General Instrument Corporation | Motion vector prediction design simplification |
Family Cites Families (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
TWI273832B (en) * | 2002-04-26 | 2007-02-11 | Ntt Docomo Inc | Image encoding device, image decoding device, image encoding method, image decoding method, image decoding program and image decoding program |
KR100946790B1 (ko) * | 2005-01-07 | 2010-03-11 | 니폰덴신뎅와 가부시키가이샤 | 영상 부호화 방법 및 장치, 영상 복호 방법 및 장치, 그들 프로그램을 기록한 기록 매체 |
KR101364195B1 (ko) * | 2008-06-26 | 2014-02-21 | 에스케이텔레콤 주식회사 | 움직임벡터 부호화/복호화 방법 및 그 장치 |
CN101677400B (zh) * | 2008-09-19 | 2012-08-15 | 华为技术有限公司 | 编码、解码方法和编码器、解码器及编解码系统 |
CN101742278B (zh) * | 2008-11-12 | 2012-11-07 | 富士通半导体股份有限公司 | 获取图像的运动矢量和边界强度的方法和系统 |
CN101860754B (zh) * | 2009-12-16 | 2013-11-13 | 香港应用科技研究院有限公司 | 运动矢量编码和解码的方法和装置 |
-
2013
- 2013-12-13 CN CN201811571925.6A patent/CN111200736B/zh active Active
- 2013-12-13 CN CN201310684728.6A patent/CN104244002B/zh active Active
-
2014
- 2014-06-13 SG SG11201510258PA patent/SG11201510258PA/en unknown
- 2014-06-13 KR KR20140072304A patent/KR20140146541A/ko not_active Application Discontinuation
- 2014-06-13 WO PCT/KR2014/005226 patent/WO2014200313A1/ko active Application Filing
- 2014-06-13 US US14/898,291 patent/US10051282B2/en not_active Expired - Fee Related
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20060153300A1 (en) * | 2005-01-12 | 2006-07-13 | Nokia Corporation | Method and system for motion vector prediction in scalable video coding |
KR20120026092A (ko) * | 2009-06-26 | 2012-03-16 | 후아웨이 테크놀러지 컴퍼니 리미티드 | 비디오 이미지의 움직임 정보를 취득하는 방법, 장치, 및 기기, 그리고 템플릿 구성 방법 |
US20110080954A1 (en) * | 2009-10-01 | 2011-04-07 | Bossen Frank J | Motion vector prediction in video coding |
US20120128060A1 (en) * | 2010-11-23 | 2012-05-24 | Mediatek Inc. | Method and Apparatus of Spatial Motion Vector Prediction |
US20130003851A1 (en) * | 2011-07-01 | 2013-01-03 | General Instrument Corporation | Motion vector prediction design simplification |
Also Published As
Publication number | Publication date |
---|---|
US20160134886A1 (en) | 2016-05-12 |
CN104244002A (zh) | 2014-12-24 |
CN104244002B (zh) | 2019-02-05 |
US10051282B2 (en) | 2018-08-14 |
SG11201510258PA (en) | 2016-01-28 |
CN111200736A (zh) | 2020-05-26 |
CN111200736B (zh) | 2022-06-21 |
KR20140146541A (ko) | 2014-12-26 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2014200313A1 (ko) | 움직임 정보를 획득하기 위한 방법 | |
WO2011087321A2 (en) | Method and apparatus for encoding and decoding motion vector | |
WO2014007551A1 (en) | Method of multi-view video sequence coding/decoding based on adaptive local correction of illumination of reference frames without transmission of additional parameters (variants) | |
WO2011090314A2 (en) | Method and apparatus for encoding and decoding motion vector based on reduced motion vector predictor candidates | |
WO2013109092A1 (ko) | 영상 부호화 및 복호화 방법 및 장치 | |
WO2021025451A1 (ko) | 움직임 정보 후보를 이용한 영상 부호화/복호화 방법, 장치 및 비트스트림을 전송하는 방법 | |
WO2017171107A1 (ko) | 인터 예측 모드 기반 영상 처리 방법 및 이를 위한 장치 | |
WO2018066927A1 (ko) | 인터 예측 모드 기반 영상 처리 방법 및 이를 위한 장치 | |
WO2017105097A1 (ko) | 머지 후보 리스트를 이용한 비디오 복호화 방법 및 비디오 복호화 장치 | |
WO2014171713A1 (ko) | 인트라 예측을 이용한 비디오 부호화/복호화 방법 및 장치 | |
WO2013002557A2 (ko) | 움직임 정보의 부호화 방법 및 장치, 그 복호화 방법 및 장치 | |
WO2017065525A2 (ko) | 영상을 부호화 또는 복호화하는 방법 및 장치 | |
WO2011126309A2 (ko) | 비디오 부호화 방법 및 장치, 비디오 복호화 방법 및 장치 | |
WO2018105757A1 (ko) | 영상 코딩 시스템에서 영상 디코딩 방법 및 장치 | |
WO2019190211A1 (ko) | 모션 보상을 이용하는 비디오 신호 처리 방법 및 장치 | |
WO2012173415A2 (ko) | 움직임 정보의 부호화 방법 및 장치, 그 복호화 방법 및 장치 | |
WO2013002589A2 (ko) | 휘도 성분 영상을 이용한 색차 성분 영상의 예측 방법 및 예측 장치 | |
WO2009110721A2 (ko) | 복수 참조에 의한 움직임 예측을 이용한 부호화 방법 및 장치, 그리고 복수 참조에 의한 움직임 예측을 이용한 복호화 방법 및 장치 | |
WO2019045427A1 (ko) | 양방향 옵티컬 플로우를 이용한 움직임 보상 방법 및 장치 | |
WO2014107074A1 (ko) | 스케일러블 비디오의 부호화 및 복호화를 위한 움직임 보상 방법 및 장치 | |
WO2013002450A1 (ko) | 보간 방법 및 이를 이용한 예측 방법 | |
WO2014003421A1 (ko) | 비디오 부호화 및 복호화를 위한 방법 | |
WO2019194568A1 (ko) | 어파인 모델 기반의 영상 부호화/복호화 방법 및 장치 | |
WO2012005549A2 (en) | Methods and apparatuses for encoding and decoding motion vector | |
WO2021054676A1 (ko) | Prof를 수행하는 영상 부호화/복호화 방법, 장치 및 비트스트림을 전송하는 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: 14811597 Country of ref document: EP Kind code of ref document: A1 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
WWE | Wipo information: entry into national phase |
Ref document number: 14898291 Country of ref document: US |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 14811597 Country of ref document: EP Kind code of ref document: A1 |
|
REG | Reference to national code |
Ref country code: BR Ref legal event code: B01A Ref document number: 112015031287 Country of ref document: BR |
|
ENP | Entry into the national phase |
Ref document number: 112015031287 Country of ref document: BR Kind code of ref document: A2 Effective date: 20151214 |