EP2561679A1 - Procédé de traitement d'une information de mouvement et procédés de codage - Google Patents
Procédé de traitement d'une information de mouvement et procédés de codageInfo
- Publication number
- EP2561679A1 EP2561679A1 EP11731021A EP11731021A EP2561679A1 EP 2561679 A1 EP2561679 A1 EP 2561679A1 EP 11731021 A EP11731021 A EP 11731021A EP 11731021 A EP11731021 A EP 11731021A EP 2561679 A1 EP2561679 A1 EP 2561679A1
- Authority
- EP
- European Patent Office
- Prior art keywords
- motion information
- information
- list
- association
- coded
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Ceased
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/50—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
- H04N19/503—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding involving temporal prediction
- H04N19/51—Motion estimation or motion compensation
- H04N19/513—Processing of motion vectors
- H04N19/521—Processing of motion vectors for estimating the reliability of the determined motion vectors or motion vector field, e.g. for smoothing the motion vector field or for correcting motion vectors
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/50—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
- H04N19/503—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding involving temporal prediction
- H04N19/51—Motion estimation or motion compensation
- H04N19/513—Processing of motion vectors
- H04N19/517—Processing of motion vectors by encoding
- H04N19/52—Processing of motion vectors by encoding by predictive encoding
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/50—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding
- H04N19/503—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using predictive coding involving temporal prediction
- H04N19/51—Motion estimation or motion compensation
- H04N19/567—Motion estimation based on rate distortion criteria
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N19/00—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- H04N19/60—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using transform coding
- H04N19/61—Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using transform coding in combination with predictive coding
Definitions
- the field of the invention is that of coding and decoding images, and in particular a video stream consisting of a series of successive images.
- the invention relates to prediction coding techniques implementing the coding of motion information relating to an image, for example the coding of motion vectors, or motion residues.
- the invention can notably apply to video coding implemented in current (MPEG, H.264, etc.) or future (ITU-T / VCEG (H.265) or ISO / MPEG (HVC) video encoders) .
- a block consists of a set of points or pixels, the block may have a square shape, rectangular, triangle or other geometric shapes.
- the coder calculates, for a current block to be encoded, a motion residual, also called motion residue, by subtracting a predictor p from the current motion vector mv:
- the coders present in H.264 adapt to the context, that is to say that the current residue can be coded differently depending on the neighboring vectors (the code words are different), they do not make it possible to consider a lower-cost objective of encoding motion information when calculating the motion information.
- the invention proposes a new solution that does not have all of these disadvantages of the prior art, in the form of a method of processing at least one motion information / mv relating to a current image of a sequence of images.
- such a method comprises:
- the invention is based on a new and inventive approach to improving encoding an image, for associating motion information relating to said image with motion information to be encoded from potential motion information.
- this association is first performed according to predetermined criteria and to achieve a coding improvement objective, expressed according to a first rule, and a goal that all the information of potential movements are associated with a single motion information to be encoded.
- the method according to the invention constructs a list comprising potential movement information, that is to say potentially able to be selected, according to predefined conventional coding criteria, for coding the image in question. .
- This list includes in particular already encoded motion information, that is to say already selected by an encoder, according to predetermined criteria, for example when coding images preceding the current image, or when coding blocks of the current image, preceding the encoding of the current block.
- This movement information is subsequently noted as "expected" motion information.
- This list also contains “theoretical" motion information, that is to say potentially usable for coding the image or block, but not previously coded, so not previously selected by an encoder, nor when encoding images preceding the current image, nor when encoding blocks of the current image, preceding the encoding of the current block.
- This movement information is subsequently noted as “unexpected" motion information.
- This potential motion information has different component values.
- the motion information is information conventionally allowing the coding of an image by motion compensation prediction from at least one reference image.
- the motion information is for example motion vectors, or movement residues, corresponding to a difference between a motion vector and a predictor (the latter corresponding for example to a vector median movement motion vectors of neighboring blocks of the current block , already coded).
- the list is modified, according to a first rule, by adding at least one association between one of the expected movement information it initially contains, and a motion information to be coded, identified among the Unexpected movement information from the start list and the expected movement information from the start list already associated during a previous iteration of this first association step.
- This motion information to be encoded is identified to potentially enhance the coding of the image or block of the image. Indeed, if the associated motion information is selected for this block, when encoding (according to predefined conventional coding criteria), this The motion information to be coded is the one that will actually be encoded for the block.
- the expected motion information is processed according to a predetermined order, defined by a processing priority criterion detailed below.
- the list is modified, according to a second rule, by adding at least one association between one of the unexpected motion information that was not associated during the first association step, and a motion information to be encoded, identified among the movement information of the start list so that all the movement information of the start list is associated once to a single motion information to be encoded.
- the list is constructed, or updated, for each block to be encoded of the current image.
- said first association step comprises a first selection step, according to at least one first processing priority criterion, of at least one of said expected movement information of said list, called current motion information, and for said current motion information, at least one iteration of the following substeps, as long as said first rule is not checked:
- a first rule is thus used to obtain motion information to be coded during the first association step.
- the first rule defines a coding improvement objective aimed at reducing, at the time of coding a block or an image, the coding cost of a movement information for this block, or this image. .
- the first rule also defines a first processing priority criterion for determining the current motion information to be considered among the expected motion information, and a second processing priority criterion for determining the unexpected motion information to be considered.
- the unexpected motion information and the expected motion information already associated are considered indiscriminately, that is, first, the unexpected motion information is not considered, and then the expected motion information already associated, but all this information one after the other as long as the objective of improving the coding cost is not achieved,
- the unexpected motion information is chosen if it can be considered that its coding cost would be a priori lower than the coding cost of the current motion information.
- said current movement information is associated with itself.
- these coding costs are calculated in order to verify the coding cost optimization hypothesis.
- said second association step comprises at least one iteration of the following steps:
- a second substep of identification according to at least a fourth processing priority criterion, of said unmatched non-associated motion information during said first association step and said associated expected motion information during said first step; of association, of a movement information to be coded and
- the second rule is that all movement information in the start list is associated with motion information at. encode, while attempting to achieve a coding cost optimization objective, for example, to obtain the least "unfavorable" coding cost of the motion information to be encoded, with respect to its associated motion information .
- the second association step also takes into account criteria (third and fourth) of processing priority of the movement information of the list, as already described above in relation to the first and second criteria.
- processing priority criteria correspond to at least one parameter belonging to the group comprising:
- the processing priority may depend on different parameters, combined or not.
- the processing priority corresponds to a processing direction of the movement information, when it is represented "graphically" in a repository.
- a repository represents motion information according to two dimensions (x, y), each corresponding to a component of the motion information, when it is representative of a movement in two dimensions.
- a frame of reference represents a movement information in three dimensions (x, y, z), each corresponding to a component of the motion information, when this is representative of a movement in three dimensions.
- motion information can be represented by a point, or a position, whose coordinates in this frame of reference correspond to the values of the components of the motion information.
- the center of the repository can for example represent a zero movement.
- the processing priority takes into account a frequency of appearance of the motion information in the previous codings performed. (coding of the images preceding the current image and / or coding of the blocks preceding the current block).
- an expected motion information may be associated with an occurrence frequency indicating the number of times it has been previously coded, i.e. selected by the coder, or an order of magnitude of the number of times it has been previously coded. For example, it may be considered that the motion information with the highest frequency of occurrence will be processed in priority.
- the processing priority takes into account a degree of confidence associated with motion information, that is, the processing priority takes into account a more or less significant probability value that a movement information of the list is effectively selected for block coding according to predefined conventional coding criteria.
- the motion information with the highest degree of confidence is treated first.
- these parameters may be combined to define the processing priority.
- processing direction belongs to the group comprising:
- processing direction is centered on a predetermined motion information or a barycenter of said motion information of said list.
- a direction of processing the information when represented in a repository, consists in traversing the points or positions representing the movement information, according to a predefined geometric template, such as a spiral or a diamond.
- a spiral is scanned away from the center thereof, this center being represented by a predetermined movement information or a barycenter of the motion information represented.
- a direction of information processing can take into account the respective positions of the represented motion information. For example, priority is given to the motion information whose positions are farthest from the center of the reference frame, that is to say the movement information representative of a large movement.
- said construction step takes into account a number n of reference images.
- the temporal activity of the scene can be taken into account. For example, if the successive images are very different from each other, then the already encoded motion information may not be relevant to the current image. In this case, it is best not to consider a large number of previous images.
- the invention also relates to a method of encoding at least one current block of a current image of an image sequence.
- such a coding method implements a treatment method as described above and the following steps:
- the invention makes it possible to code motion information in an optimized manner.
- the coded motion information is not directly the selected motion information, according to conventional criteria predetermined for coding, but a modified motion information according to the invention, associated with the selected motion information.
- the coding of a current image is optimized because it has a globally reduced coding cost for all the blocks of the image, compared with the prior art.
- the reduction of the coding cost corresponds to a reduction in the cost of transport, between an encoder and a decoder, of the movement information associated with the blocks of an image.
- the transmission rate can be increased compared with the prior art.
- Another aspect of the invention relates to a signal representative of at least one coded block of an image of a sequence of images.
- a signal carries at least one encoded motion information according to the coding method described above.
- such a signal also carries information representative of a number n of reference images.
- the invention also relates to a method for decoding at least one current block of a current image of a sequence of images.
- such a decoding method comprises the following steps: receiving and decoding a coded motion information according to the encoding method described above, delivering decoded motion information;
- Another aspect of the invention relates to a device for processing at least one movement information relating to a current image of a sequence of images.
- a treatment device comprises:
- Means for constructing at least one list L comprising at least one motion information, called expected motion information iinvi-a, resulting from a motion compensation prediction from at least one reference image and having has been previously coded, and at least one motion information, called untimely movement information Imvi- ⁇ , not previously coded;
- said first association means being adapted to respect a first coding cost optimization rule of said motion information to be coded and to issue a modified list L comprising at least one association (Imv r a, ⁇ );
- said second association means being adapted to comply with a second rule that each of the movement information Jmvi of said list is associated, by said first or second association means, with a single motion information to be encoded ⁇ , the second association means delivering a modified list L including as many associations (Imv i -a, I'mv i ) as motion information in the list L constructed by the means of construction.
- Such a processing device is particularly suitable for implementing the steps of the treatment method described above.
- This device may of course include the various features relating to the treatment method according to the invention.
- the features and advantages of this treatment device are the same as those of the treatment method, and are not detailed further.
- a coding device also called coder
- a coding device of at least one current block of a current image of a sequence of images.
- a coding device comprises means of the treatment device described above and:
- the invention also relates to a device for decoding at least one current block of a current image of a sequence of images, comprising:
- Means for reconstructing said block from said selected motion information Means for reconstructing said block from said selected motion information.
- the invention relates to a computer program comprising instructions for the implementation of a method of processing, coding or decoding as described above, when this program is executed by a processor.
- FIG. 1 already commented on in relation with the prior art, illustrates an exemplary encoding system according to the prior art H.264 technique;
- FIG. 2 illustrates the main steps of the method of processing an image-related motion information according to an embodiment of the invention
- FIGS 3a and 3b respectively illustrate the first and second association steps. the treatment method of Figure 2;
- FIGS. 4a, 4b, 5a, 5b and 6 illustrate examples of representation of motion information for a current block of an image
- FIG. 7 illustrates an example of a simplified structure of a processing device, according to one embodiment of the invention.
- the general principle of the invention is based on the processing of a motion information Imv relating to an image of a sequence of images, so as to deliver to the encoder, for the coding of the image, or a block of the image, a modified motion information I'mv, hereinafter called motion information to be encoded, whose coding cost is a priori optimized with respect to the motion information before processing.
- the invention makes it possible to obtain an overall improvement in the encoding cost of a sequence of video images, by not directly coding the motion information Imv relative to each image, but a motion information processed I'mv according to the invention.
- the invention is based on a set of potentially useful movement information for coding the block, or the image.
- the construction of this assembly according to the invention then allows the processing of each of the motion information that it contains, so that the motion information that will actually be selected for the coding of the block, or of the image, is processed. also.
- Motion information Imv is considered to be information conventionally enabling the coding of an image by motion compensation prediction from at least one reference image.
- a motion information is, for example, a motion vector, or a movement residue, corresponding to a difference between a motion vector and a predictor (the latter corresponding for example to a vector median movement of motion vectors of the neighboring blocks of the current block , already coded), or a predictor.
- the processing of the motion information according to the principle of the invention is also implemented, in order to efficiently decode the motion information relative to the block, or image, in the process of reconstruction.
- a motion information Imv corresponds to a movement residue Bnv, calculated for example according to the known technique described in relation with the prior art, for a block of a current image of a sequence images.
- the treatment method according to this embodiment of the invention mainly comprises three stages 20, 21 and 22.
- a first step of building a list L all the previously selected and coded motion information is coded during the coding of one or more images preceding the current image and / or during the coding of blocks preceding the current image. . current block.
- This movement information is subsequently called expected motion information, and noted Imvi-a.
- step 20 all the potentially selectable motion information from the encoder, different from the motion information Imvi-a, called theoretical motion information or unexpected motion information, and noted Imvi -n / A.
- FIG. 4a firstly presents an "Image" image, in which a current block is being processed according to the method of the invention. This block is represented by a hatched square. Previously coded neighboring blocks are represented by white squares with black dots
- motion information mvj to mv4 is represented in the form of vectors by arrows along the x and y axes each corresponding to a component of the motion information.
- This motion information corresponds to the previously coded motion vectors of the neighboring blocks of the current block.
- the motion information mvj corresponds approximately to a movement of two blocks to the right, and two blocks to the bottom, while the motion information mvj corresponds approximately to a movement of a block to the right, and a block up.
- the motion information P corresponds to the calculated predictor for the current block.
- FIG. 4b is represented a specific reference centered on the predictor P, under the form of a dotted window, making it possible to represent relative to the predictor P of the current block the motion information Imv by points, also called positions, according to their components.
- This repository corresponds to the repository of the potential motion residuals for the coding of the motion residual of the current block.
- Each of the motion information Imv1 to Imv4 is the difference between. on the one hand respectively each motion vector mv 1 to mv 4 and, on the other hand, the motion information P (the motion vector predictor).
- the size of the dotted window being predetermined and defining the size of the list L.
- this number n of images preceding the current image taken into account for the construction of the list may vary, according to certain criteria, related for example to the temporal activity of the image sequence.
- this number n must be known to the decoder, and can for example be transmitted to it in a signal.
- the motion information from reference images and therefore temporally distant from the current image, are classically rescaled, so as to manage the temporal distance.
- this list can include the movement information of the blocks previously coded to the current block, so it can be updated for each block.
- an objective is to associate with each of the information Imv t -, including the motion information Imvj-a and Imvj-na, a unique motion information ⁇ , having a priori an optimized coding cost.
- a motion information Jmv / is selected for the current block, according to known coding criteria, it is the motion information to be coded / W; associated according to this embodiment of the invention which is encoded by the encoder, thereby optimizing the coding cost of the block.
- a first association step 21 consists of associating, when possible, all the expected motion information Imvfa of the list L with motion information I'mv i , chosen from the unexpected motion information Imv t -na from the list L and the expected motion information associated with a previous iteration of the first step 21, so that the coding cost of I'mvi is a priori lower than that of Imvi-a.
- This optimization of the coding cost is defined according to a first rule.
- This first association step may for example consist of creating a second column in the list L, or of creating links between the various elements of the list. L, thus modifying it in a list L '.
- each element of the starting list L that has not been associated during step 21 is processed, so that, at the end of this second association step 22, each movement information / mv / from the start list L is associated with a motion information to code ⁇ ,.
- the movement information processed during this second step is unexpected motion information Imv i -na that has not "served” an association, in the first step 21,
- Figure 5a graphically represents the motion information of the list L as black dots for the expected motion information hnv r - a and white for the unexpected motion information Iinvi-na.
- FIG. 3a illustrates the main substeps of the association step 21, namely a selection sub-step 210, an identification sub-step 211 and an association sub-step 212 itself.
- the first selection sub-step 210 firstly consists in selecting, according to a processing priority criterion, a motion information Imvi-a from the list L to be processed first. Since the purpose of this first association step 21 is an optimization of the coding cost, it is preferable to process Imvi-a motion information considered to have a higher potential of being selected at the time of coding. Thus, the motion information that will actually be chosen coding is well associated with a motion information to code whose coding cost is optimized.
- this first selection consists in choosing a black point, among those represented in FIG. 5a.
- the treatment priority criterion may for example be based on the occurrence frequency related to each movement information of the list, and corresponding to a piece of information. representative of the number of times a motion information has been previously coded. For example, the occurrence frequency is equal to the exact number of times the motion information has already been coded, during the coding of the n preceding images and blocks preceding the current block. The frequency of occurrence may also be equal to an order of magnitude of the number of times it has already been coded, such as "greater than 20", or "less than 5".
- This processing priority criterion thus makes it possible to process, in priority, the most frequent expected Imvi-a motion information, and therefore to associate to them motion information to code Z'mv / that is optimal in terms of coding cost.
- motion information has never been previously coded (which is the case, for example, of the unexpected information Imvi-na described above), their occurrence frequency is zero.
- the priority criterion may also be based on a geometric direction of travel when the motion information is graphically represented, as in Figure 5a.
- a sense of processing consists in taking the expected motion information in the order in which it appears when starting from the center of the repository, towards the outside of the repository, by going through a spiral (such as for example the spiral connecting the white dots in Figure 5a which is traversed in the direction indicated by the double arrows).
- the direction of travel can also be represented as a rhombus, centered on the center of the repository.
- the geometric path sense examples described above may also be centered on another point in the repository distinct from the center, for example a point representing the center of gravity of all the motion information of the list L.
- the direction of travel can also take into account a calculated distance between the point to be treated and another point considered as a reference.
- the processing priority criterion may also be based on information representative of a degree of confidence associated with the motion information of the list L.
- the motion information of the list L having a high degree of confidence can be processed by priority, which allows them to associate motion information coding / 'mv / optimal in terms of coding cost.
- the treatment priority criterion used is to number the black dots, for example from 1 to 7, and to process them in the ascending order of their number. .
- This identification is based on the first rule already described above, which consists in optimizing the coding cost of a movement information.
- an unexpected motion information or an expected motion information but already associated with an unexpected motion information during a previous iteration (for example point 6 is associated with point 2, having itself already was associated with a white point during a previous iteration), which would potentially cost less to code than the currently selected motion information, among all the unexpected motion information Imv ⁇ -na from the L list.
- a treatment priority criterion as previously described.
- This first association step 21 is implemented for all the expected movement information of the list, and gives rise to as many associations. It can therefore be considered that the list L has been modified following this first step 21 into a list ⁇
- FIG. 5b A representation of this modified list L is presented in FIG. 5b.
- Previous blackheads that have been associated with white dots have become white with black dots, and the associated white dots have become black dots with white dots.
- the positions corresponding to the black dots of Figure 5a are "released” in Figure 5b.
- the positions corresponding to the white dots of FIG. 5a are now "occupied” in FIG. 5b.
- Step 22 also comprises three substeps 220, 221 and 222, identical in principle to the sub-steps 210, 211 and 212 described above, in relation to the first association step 21. Indeed, substeps 220, 221 and 222 differ from substeps 210, 21 1 and 212 only by the motion information they process. This difference is essentially related to the fact that the objective of this second association step is that each movement information of the list L is associated with a unique motion information to be encoded I'mv.
- the three selection sub-steps 220, identification 221 and association 222 are also based on processing priority criteria, as previously described.
- the selection sub-step 220 takes into account all the unexpected non-associated motion information in the first association step, i.e. the motion information graphically represented in FIG. 5a by white dots ;
- the identification sub-step 221 takes into account the motion information graphically represented in FIG. 5a by white dots and white dots with black dots, so that the motion information to be coded for the identified mv has a priori an optimal coding cost, or the least unfavorable;
- association sub-step 222 makes it possible to associate with each movement information selected in the substep 220 a unique motion information to be coded I'mv identified during the sub-step 221. These associations are represented on the Figure 5b by dashed lines.
- information from The motion to be coded I'mv may have a priori higher coding cost, but as least unfavorable as possible, than the selected motion information to which this motion information to be encoded I'mv is associated.
- this second association step does not deal with the most frequent expected motion information (as indicated above in relation to the processing priority criteria), this has no overall effect on the coding cost. of the image.
- FIG. 6 illustrates a variant of the embodiment described above, in which only part of the list L constructed, called sub-list L ', is considered in order to compromise between coding quality and complexity of implementation. .
- a first type of restriction applied to the list L may be based on the type of expected motion information it contains. Indeed, it is preferable for example the expected motion information from the previous coding of neighboring blocks in the image of the current block.
- the sublist 1 then corresponds to the "sub-window" 1 illustrated in Figure 6, dotted.
- a second type of restriction applied to the list L may be based on a degree of confidence associated with the expected motion information it contains. Indeed, it is preferable for example the expected motion information having the highest probability of being effectively selected coding.
- Sub-list 2 then corresponds to the "sub-window" 2 illustrated in Figure 6, in solid line.
- the steps of the processing method described above are implemented by a method of encoding a block of an image.
- the coding method selects, from the motion information of the list L, motion information corresponding to conventional predefined criteria for coding, such as for example the "rate-distortion” criterion. ".
- the "rate” parameter of the "rate-distortion” criterion is then calculated from the motion information to code the associated I'mv.
- the motion information actually coded for the block, and transmitted for to be decoded by a decoder at the time of the reconstruction of the block of the image is the motion information to be coded I'mv which has been associated with it during the steps of the treatment method described above, and which presents a priori a lower coding cost.
- the coding cost of the block of the image is optimized, while ensuring optimal coding performance.
- the motion information I'mv transmitted by the encoder is considered as motion information to be decoded.
- the decoder When decoding a block, the decoder implements the same steps as those implemented by the coder, that is to say that it implements the steps of the processing method, so as to obtain a list and a plurality of associations between motion information Imv and the motion information to be decoded I'mv.
- the decoder decodes the received information, transmitted by the coder, then identifies in the list, among the motion information to be decoded I'mv, that which corresponds to the motion information that has just decoded. To find the motion information corresponding to the block he is rebuilding, he looks for the associated motion information Imv.
- FIG. 7 illustrates an example of a simplified structure of a processing device according to one embodiment of the invention.
- the processing device comprises a memory 71 consisting of a buffer memory, a processing unit 72, equipped for example with a microprocessor ⁇ , and driven by the computer program 73, implementing the processing method according to the invention.
- the code instructions of the computer program 73 are for example loaded into a RAM memory before being executed by the processor of the processing unit
- the processing unit 72 receives as input at least one relative motion information, an image of a sequence of images.
- the microprocessor of the processing unit 72 implements the steps of the processing method described above, according to the instructions of the computer program 73, to output at least one motion information to be encoded.
- the processing device comprises, in addition to the buffer memory 71:
- a module M20 for constructing at least one list L comprising at least one motion information, called the expected motion information Imvi-a, resulting from a prediction by motion compensation from at least one image of reference and having been previously coded, and at least one motion information, called unexpected motion information Imv [-na, not previously coded.
- a second module M22 associating each unexpected motion information not associated with the first association module with a motion information to be coded ⁇ ; identified among the expected non-associated motion information by the first association module and the expected motion information associated with the first association module, the second association module being adapted to comply with a second rule that each Imvi motion information of the list is associated, by the first or second association modules, with a single motion information to be encoded I'mvi, said second association module delivering a modified list L comprising as many associations ⁇ Imvi-a, I'mvi) motion information in the list L constructed by the building module M20.
- These modules are driven by the microprocessor of the processing unit 72.
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Compression Or Coding Systems Of Tv Signals (AREA)
Abstract
Description
Claims
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
FR1053060A FR2959376A1 (fr) | 2010-04-22 | 2010-04-22 | Procede de traitement d'une information de mouvement, procedes de codage et de decodage, dispositifs, signal et programme d'ordinateur correspondants |
PCT/FR2011/050898 WO2011131903A1 (fr) | 2010-04-22 | 2011-04-19 | Procede de traitement d'une information de mouvement et procedes de codage |
Publications (1)
Publication Number | Publication Date |
---|---|
EP2561679A1 true EP2561679A1 (fr) | 2013-02-27 |
Family
ID=43127333
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
EP11731021A Ceased EP2561679A1 (fr) | 2010-04-22 | 2011-04-19 | Procédé de traitement d'une information de mouvement et procédés de codage |
Country Status (5)
Country | Link |
---|---|
US (1) | US9560376B2 (fr) |
EP (1) | EP2561679A1 (fr) |
CN (1) | CN102959954B (fr) |
FR (1) | FR2959376A1 (fr) |
WO (1) | WO2011131903A1 (fr) |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
FR2959376A1 (fr) * | 2010-04-22 | 2011-10-28 | France Telecom | Procede de traitement d'une information de mouvement, procedes de codage et de decodage, dispositifs, signal et programme d'ordinateur correspondants |
CN107277506B (zh) * | 2017-08-15 | 2019-12-03 | 中南大学 | 基于自适应运动矢量精度的运动矢量精度选择方法及装置 |
Family Cites Families (19)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US6178265B1 (en) | 1994-09-22 | 2001-01-23 | Intel Corporation | Method and apparatus for motion vector compression |
JP4102973B2 (ja) * | 2002-04-24 | 2008-06-18 | 日本電気株式会社 | 動画像の符号化方法および復号化方法、これを用いた装置とプログラム |
US8254461B2 (en) * | 2002-07-24 | 2012-08-28 | Apple Inc. | Method and apparatus for variable accuracy inter-picture timing specification for digital video encoding with reduced requirements for division operations |
US8000392B1 (en) * | 2004-02-27 | 2011-08-16 | Vbrick Systems, Inc. | Phase correlation based motion estimation in hybrid video compression |
JPWO2007074605A1 (ja) * | 2005-12-26 | 2009-06-04 | 株式会社ニコン | 画像処理方法、画像処理プログラム、画像処理装置、及び撮像装置 |
US7672377B2 (en) * | 2006-04-21 | 2010-03-02 | Dilithium Holdings, Inc. | Method and system for video encoding and transcoding |
KR101311403B1 (ko) * | 2006-07-04 | 2013-09-25 | 삼성전자주식회사 | 영상의 부호화 방법 및 장치, 복호화 방법 및 장치 |
DE102006043707A1 (de) * | 2006-09-18 | 2008-03-27 | Robert Bosch Gmbh | Verfahren zur Datenkompression in einer Videosequenz |
KR101366242B1 (ko) * | 2007-03-29 | 2014-02-20 | 삼성전자주식회사 | 움직임 모델 파라메터의 부호화, 복호화 방법 및 움직임모델 파라메터를 이용한 영상의 부호화, 복호화 방법 및장치 |
CN101159875B (zh) * | 2007-10-15 | 2011-10-05 | 浙江大学 | 二重预测视频编解码方法和装置 |
CN101933328B (zh) * | 2008-01-22 | 2014-11-19 | 杜比实验室特许公司 | 利用动态查询表更新的自适应运动信息成本估计 |
WO2009115901A2 (fr) * | 2008-03-19 | 2009-09-24 | Nokia Corporation | Vecteur de mouvement combiné et prédiction d'indice de référence pour codage vidéo |
US20090316042A1 (en) * | 2008-06-19 | 2009-12-24 | Chin-Chuan Liang | Video processing methods and related devices for determining target motion vector |
EP2227012A1 (fr) * | 2009-03-05 | 2010-09-08 | Sony Corporation | Système et procédé pour la fourniture de vecteurs de mouvement fiables |
US8873626B2 (en) * | 2009-07-02 | 2014-10-28 | Qualcomm Incorporated | Template matching for video coding |
FR2959376A1 (fr) * | 2010-04-22 | 2011-10-28 | France Telecom | Procede de traitement d'une information de mouvement, procedes de codage et de decodage, dispositifs, signal et programme d'ordinateur correspondants |
US9300970B2 (en) * | 2010-07-09 | 2016-03-29 | Samsung Electronics Co., Ltd. | Methods and apparatuses for encoding and decoding motion vector |
JP5298140B2 (ja) * | 2011-01-12 | 2013-09-25 | 株式会社エヌ・ティ・ティ・ドコモ | 画像予測符号化装置、画像予測符号化方法、画像予測符号化プログラム、画像予測復号装置、画像予測復号方法、及び画像予測復号プログラム |
EP2787728B1 (fr) * | 2011-09-23 | 2020-11-11 | KT Corporation | Procédé de provocation d'un bloc candidat à la fusion |
-
2010
- 2010-04-22 FR FR1053060A patent/FR2959376A1/fr not_active Withdrawn
-
2011
- 2011-04-19 US US13/642,779 patent/US9560376B2/en active Active
- 2011-04-19 CN CN201180019807.4A patent/CN102959954B/zh active Active
- 2011-04-19 EP EP11731021A patent/EP2561679A1/fr not_active Ceased
- 2011-04-19 WO PCT/FR2011/050898 patent/WO2011131903A1/fr active Application Filing
Non-Patent Citations (2)
Title |
---|
JULIEN LE TANOU ET AL: "Motion vector forecast and mapping (MV-FMap) method for entropy coding based video coders", 2010 IEEE INTERNATIONAL WORKSHOP ON MULTIMEDIA SIGNAL PROCESSING (MMSP '10), SAINT MALO, FRANCE, 4-6 OCT. 2010, IEEE, IEEE, PISCATAWAY, USA, 4 October 2010 (2010-10-04), pages 206 - 211, XP031830583, ISBN: 978-1-4244-8110-1 * |
See also references of WO2011131903A1 * |
Also Published As
Publication number | Publication date |
---|---|
CN102959954A (zh) | 2013-03-06 |
US9560376B2 (en) | 2017-01-31 |
WO2011131903A1 (fr) | 2011-10-27 |
US20130039425A1 (en) | 2013-02-14 |
CN102959954B (zh) | 2016-05-11 |
FR2959376A1 (fr) | 2011-10-28 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
EP2446628B1 (fr) | Procédé de décodage d'images, dispositifs de décodage et programme d'ordinateur correspondants | |
EP2991350B1 (fr) | Procédé de décodage d'images | |
EP3061246B1 (fr) | Procédé de codage et de décodage d'images, dispositif de codage et de décodage d'images et programmes d'ordinateur correspondants | |
EP3651466A1 (fr) | Prédiction d'un vecteur mouvement d'une partition d'image courante de forme géometrique ou de taille différente de celle d'au moins une partition d'image de référence voisine, codage et decodage utilisant une telle prediction | |
EP2347590B1 (fr) | Prediction d'une image par compensation en mouvement en avant | |
FR2933565A1 (fr) | Procede et dispositif de codage d'une sequence d'images mettant en oeuvre une prediction temporelle, signal, support de donnees, procede et dispositif de decodage, et produit programme d'ordinateur correspondants | |
EP2443835B1 (fr) | Codage de vecteurs mouvement par compétition de prédicteurs | |
EP3075155B1 (fr) | Procédé de codage et de décodage d'images, dispositif de codage et de décodage d'images et programmes d'ordinateur correspondants | |
EP2392142A1 (fr) | Procedes de codage et de decodage d'une sequence d'image mettant en oeuvre une compensation en mouvement, dispositifs de codage et de decodage, signal et programmes d'ordinateur correspondants | |
EP2486659B1 (fr) | Procedes de codage et de décodage d'images, dispositifs de codage et de decodage et programmes d'ordinateur correspondants | |
WO2011131903A1 (fr) | Procede de traitement d'une information de mouvement et procedes de codage | |
FR3040578A1 (fr) | Procede de codage et de decodage d'images, dispositif de codage et de decodage d'images et programmes d'ordinateur correspondants | |
EP3360328B1 (fr) | Codage et décodage multi-vues | |
WO2020070409A1 (fr) | Codage et décodage d'une vidéo omnidirectionnelle | |
WO2019115899A1 (fr) | Procédés et dispositifs de codage et de décodage d'une séquence vidéo multi-vues représentative d'une vidéo omnidirectionnelle | |
WO2010086562A1 (fr) | Procede et dispositif de codage d'images mettant en oeuvre des modes de codage distincts, procede et dispositif de decodage, et programmes d'ordinateur correspondants | |
WO2018065687A1 (fr) | Procede de codage d'une image, procede de decodage, dispositifs, equipement terminal et programmes d'ordinateurs associes | |
FR2957744A1 (fr) | Procede de traitement d'une sequence video et dispositif associe | |
WO2024042286A1 (fr) | Lissage hors boucle de codage d'une frontière entre deux zones d'image | |
FR2956552A1 (fr) | Procede de codage ou de decodage d'une sequence video, dispositifs associes | |
WO2020058595A1 (fr) | Procédés et dispositifs de codage et de décodage d'un flux de données représentatif d'au moins une image | |
WO2011131902A2 (fr) | Procédé d'enrichissement d'une information de mouvement et procédé de codage |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PUAI | Public reference made under article 153(3) epc to a published international application that has entered the european phase |
Free format text: ORIGINAL CODE: 0009012 |
|
17P | Request for examination filed |
Effective date: 20121113 |
|
AK | Designated contracting states |
Kind code of ref document: A1 Designated state(s): AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR |
|
DAX | Request for extension of the european patent (deleted) | ||
RAP1 | Party data changed (applicant data changed or rights of an application transferred) |
Owner name: ORANGE |
|
17Q | First examination report despatched |
Effective date: 20170606 |
|
REG | Reference to a national code |
Ref country code: DE Ref legal event code: R003 |
|
STAA | Information on the status of an ep patent application or granted ep patent |
Free format text: STATUS: THE APPLICATION HAS BEEN REFUSED |
|
18R | Application refused |
Effective date: 20180322 |