Abstract: The present invention relates to a device for compressing a source image by means of a base layer, coding a base image, and of an enhancement layer, coding a residual existing between the source image and the base image, the source image and the base image being structured as blocks of pixels. According to the invention, such a device is characterized in that it comprises means for generating the enhancement layer of a block as a function of the quality of restitution of this block, determined by comparing the coding of this block in the source image with the coding of this block in the base image, and/or as a function of information coded by this block of pixels.

Abstract: A method for decoding a stream of coded data representative of a sequence of images is disclosed. The decoding method comprises: determining a prediction block, calculating the average of said prediction block, reconstructing, from said stream, a DC coefficient and AC coefficients relating to said current block, calculating an average of said current block from said average of said prediction block and said DC coefficient, calculating a weighted prediction parameter as being the ratio of the average of said current block and the average of said prediction block, transforming a block of coefficients comprising a null value as DC coefficient and said reconstructed AC coefficients into a residual block, reconstructing the current block by merging said residual block and said prediction block weighted by said weighted prediction parameter. A method for coding is further disclosed.

Abstract: Super-resolution refers to a process of recovering the missing high-frequency details of a given low-resolution image. Known single image SR algorithms are often computationally intractable or unusable for most of the practical applications. The invention relates to a method for performing hierarchical super-resolution based on self content neighboring patches information is based on pyramidal decomposition. The intrinsic geometric property of an input LR patch neighborhood is obtained from the input LR patch and its K nearest neighbors taken from different down-scaled versions of the LR image.

Abstract: The invention relates to image block encoding and decoding. More precisely, the invention relates to block prediction using reference blocks of the image. It is proposed a method which comprises using processing means for using at least the block for determining one reference block in an already encoded part of the image, encoding and reconstructing a part of the image, using at least the block for determining one reference block in the reconstructed part and a vector referencing the one reference block, using the one reference block for determining at least one further reference block in the reconstructed part and encoding the determined vector and a residual of the block with respect to a combination of the one and the at least one further reference block. Using the block for determination of the one reference block allows for selecting it such that it resembles the block and correlation with the block is maintained.

Abstract: The invention relates to a method for predicting a pixel block of an image using a weighted sum of pixel blocks belonging to patches of a dictionary from a set of candidate patches, each patch being formed of a pixel block of an image and a causal neighborhood of this pixel block. The method is wherein a subset of candidate patches is obtained from the set of candidate patches and the dictionary is formed of a patch of the subset of candidate patches, called the first patch, and at least one other patch of said set of candidate patches, called the second patch.

Abstract: The process comprising splitting of an image into blocks, intra coding of a current block using spatial prediction based on a matching pursuit algorithm selecting, from a dictionary of atoms, the atom the most correlated with a causal neighborhood of the current block, is characterized in that it performs the following steps: — determination of the two dimensional shift between the causal neighborhood and the selected atom, — generation of at least, a new phased atom taking into account the values of the two dimensional spatial shift, — use of this new atom for intra prediction, according to the matching pursuit algorithm, if better correlated than the selected one. Application to video data compression.

Abstract: A method for compressing video operates by first deleting one or more video packets from binary train to provide a reduced binary train. Thereafter, error masking is performed on the reduced binary train to yield a corrected degraded local decoded image which is the image reconstructed on the basis of the reduced binary train and of the masking algorithm. The reduction in the binary train is validated as a function of the quality of the corrected degraded local decoded image. Lastly one of the reduced binary train or of the binary train, is selected depending on whether or not there is validation.

Abstract: The invention is related to a method for determining a filter for interpolating one or more pixels comprised in a frame. The proposed method comprises the step of determining the filter based on other pixels in the remainder of the frame wherein a motion vector associated with said pixels has at least one motion parameter in common with another motion vector associated with said other pixels and the filter is determined such that a quality of an interpolation of said other pixels, said interpolation being determined by help of said determined filter, reaches or exceeds the quality of any other interpolation of said other pixels determinable by help of one of different filter candidates comprised in a filter candidate set. In determining the filter solely on different other pixels of the same frame, said method for determining a filter for interpolation allows for realization of a single pass encoding framework.

Abstract: A method for reconstruction of a picture sequence coded in accordance with a coding method specifying a set of coding tools and/or their associated coding parameters is disclosed. The pictures being divided into coding entities. The method for reconstruction according to the invention comprises the following steps for each coding entity coded in INTER mode: determining for the coding entity at least one reference picture, and reconstructing the coding entity from the at least one reference picture with coding tools configured by coding parameters associated with the coding tools. Advantageously, the coding tools and/or the associated coding parameters depend on the reference picture.

Abstract: A method for coding a current block is disclosed. The method comprises an iterative step of decomposition into atoms of a first dictionary of a vector of data comprising at least reconstructed image data associated with pixels situated in a causal neighboring area of a current block. At each iteration, the step of decomposition into atoms comprises the following steps for: selecting a first atom in the first dictionary of atoms, applying a phase correlation between a patch associated with the first atom and a current residue vector, when the first atom is a textured atom, extracting from the patch at least the part most correlated with the residue vector, the extracted part forming a second atom, selecting from among the first atom and the second atom, the atom most correlated with the current residue vector, and updating the residue vector according to the selected atom.

Abstract: According to the compressed video data stream switching method for the transmission of a first data stream corresponding to a first video sequence, followed by, in a given switching point of this first stream, a second data stream corresponding to a second video sequence, the second data stream is modified prior to transmission, by a replacement of the bidirectional pictures with bidirectional commutation pictures of which the pictures blocks have a null residue and a null motion vector which is either forward or backward, according to the position of the blocks in the picture.

Abstract: The method comprises a coding of a source block according to a coding mode wherein the selection of the coding mode is carried out, for a source block, according to coding error estimations and source block coding cost estimations for the different coding modes tested, the coding error being estimated, for a coding mode, according to the sum of the absolute values of coefficients obtained after a mathematical transformation in the frequency domain of the residue block related to this coding mode and as a function of the sum of absolute values of quantized coefficients corresponding to this coding mode. An application is the decision of the coding mode for data using the MPEG 2 or MPEG4-AVC standard.

Abstract: The method wherein the field or frame mode is selected according to the following steps: determination of a motion vector associated with a co-located macroblock of a macroblock to be coded and finding in the next reference picture, a predefined macroblock to be coded, for the selection, in field or frame mode, scaling of the motion vector according to the temporal distances between the reference pictures corresponding to this motion vector and between the current picture, field or frame, according to a predefined mode and a reference picture chosen from among the reference pictures corresponding to this motion vector, determination of the temporal macroblock in the chosen reference picture, designated by the motion vector scale attributed to the macroblock to be coded, and its field or frame coding mode, selection of field or frame coding mode of the temporal macroblock for the coding of the current macroblock.

Abstract: A method for arithmetic coding of symbols in a stream is described. The method comprises the following steps: coding a current symbol with a current probability model, and updating the current probability model according to the coding of the current symbol. selecting the current probability model in a set of at least two probability models according to a coding cost criterion, and coding an identifier of the selected probability model.

Abstract: A method for reconstruction of a current block predicted from at least one reference block is described. The method comprises the steps for: decoding at least one weighted prediction parameter, called an explicit parameter, from coded data, calculating at least one weighted prediction parameter, called an implicit parameter, from a first causal neighbouring area of the current block and a first corresponding neighbouring area of the reference block, calculating a first distortion resulting from the prediction of a second causal neighbouring area of the current block with the explicit parameter and a second distortion resulting from the prediction of said second causal neighbouring area with the implicit parameter, comparing the first and second distortions, selecting a parameter between the explicit parameter and the implicit parameter according to the result of the comparison, and reconstructing the current block using the selected parameter.

Abstract: A method for extracting an epitome from an image Y is described. The method comprises: searching for each block, a list of patches in the image Y with similar content; determining, for each patch of said lists, the set of blocks able to be represented by the patch; constructing, from the sets of blocks, epitome charts comprising, for each epitome chart, initializing said epitome chart with one patch and extending said epitome chart with at least one candidate region contained in a patch overlapping said epitome chart; said method being characterized in that said candidate region is selected by taking into account the blocks represented by said patch overlapping said epitome chart and by any other patch contained in said epitome chart enlarged by said selected candidate region.

Abstract: A method of coding a sequence of images is disclosed. The method of coding comprises for a current image the steps of: creating a summary of said current image; encoding said summary into a first bitstream; reconstructing an intermediate image from said summary; and encoding, into a second bitstream, the current image using said intermediate image as reference image.

Abstract: A method of coding an image epitome is disclosed. The method of coding comprises: creating the epitome of the image, the epitome comprising a texture epitome and a transform map; coding the texture epitome and the transform map; wherein texture epitome is padded, before the step of coding, such that the texture epitome is block aligned.

Abstract: A method for coding a current block using a parametric motion model whose parameters are defined from the motion data of the causal neighbourhood is disclosed. Specifically, the method for coding the current block comprises the following steps: determine a prediction block of the current block from a motion model defined by a set of parameters, the model associating a motion vector pointing at a pixel of a prediction block with each pixel of the current block. calculate a residual block between the current block and the prediction block, and code the residual block. According to the invention, the parameters of the motion model are determined from the motion data associated with the causal neighbourhood blocks of the current block. Therefore, they can be deducted from parts of images already coded and do not need to be incorporated in the image data stream.

Abstract: A method for coding a block of pixels is disclosed. The method for coding comprises: calculating a block of residues from the pixel block and a prediction block, transforming the block of residues into a block of coefficients with a transform defined by a set of basis functions, coding the block of coefficients, The method comprises, before the transformation step, a step for rephasing basis functions from residues calculated in a causal neighbourhood of the pixel block. The transformation step uses the rephased basis functions.