Process for the format conversion of an image sequence
View Patent ↗A process for converting the format of an image sequence wherein, for a coded pixel group to be converted, if the mode of coding used is of the inter type with no residue, the conversion is performed by a copy of a converted pixel group of a preceding image linked by the motion vector associated with said coded pixel group.
1. Process for the format conversion of an image sequence employing video data coded on the basis of a structure of pixel groups comprising a first step for decoding the coded data and a second step of converting of the decoded data, wherein, for a coded pixel group to be converted, if the decoding mode is of the “inter” type with no residue, the conversion is performed by a copy of a converted pixel group of a preceding image linked by the motion vector associated with said coded pixel group.
2. Process according to claim 1 , wherein, if the motion vector associated with the pixel group is null, the conversion is performed by recopy of the co-located pixel group and, if the motion vector is different from zero, the conversion is performed by motion compensation in a preceding convened image.
3. Process according to claim 1 , wherein the data are coded according to the MPEG standard, the pixel group is an image block and the coding mode is determined from the cbp (coded block pattern) code defining the apportionment of the coded blocks in a macroblock.
4. Process according to claim 1 , wherein the data are coded according to the MPEG standard, the pixel group is a macroblock and said coding mode is determined from the “skipped macroblock” or “uncoded” mode.
5. Process according to claim 1 , the format conversion being supplemented with a modification of the display employing a simple mathematical operation applicable at the decoded pixel group level, wherein the operation, adapted to the display domain, is applied to the copied convened pixel group.
6. Process according to claim 5 , wherein the simple operation is the addition of an offset.
7. Process for the format conversion of an image sequence employing video data coded on the basis of a structure of pixel groups comprising a first step for decoding the coded data and a second step of converting of the decoded data, wherein, if for the decoding of a pixel group to be converted an error of transmission of the coded data brings about an error masking mode equivalent to a decoding of the inter type with no residue, the conversion is performed by a copy of a converted pixel group of a preceding image linked by the motion vector associated with said coded pixel group.