Manipulating sub-pictures of a compressed video signal
View Patent ↗The invention relates to manipulation of sub-pictures ( 303 ) in a picture such as a mosaic screen ( 301 ). An apparatus comprises a sub-picture generator ( 105 ) connected to a mosaic screen generator ( 107 ), which generates the mosaic screen ( 301 ) from the sub-pictures. The mosaic screen ( 301 ) is by a video compressor ( 111 ) divided into uncompressed picture blocks such that each block only comprise video data from one sub-picture. The picture blocks are compressed using a block compression scheme such as MPEG-2. The apparatus further comprises a video manipulation processor ( 113 ), which manipulates the sub-pictures by manipulating the association between control data in the video signal and the compressed picture blocks. The video data of the compressed picture blocks are not affected whereby de-compression and new compression is avoided.
1. A method of manipulating video sub-pictures, comprising:
generating from a single source, a plurality of sub-pictures from an uncompressed video signal;
generating a mosaic screen of the sub-pictures;
compressing the mosaic screen into compressed macro blocks using a video block based compression scheme, the compressed macro blocks containing horizontally and vertically arranged compressed sub-pictures in which each macro block contains only one respective compressed sub-picture; and
manipulating a horizontal location of a first compressed sub-picture by associating macro block control data of a compressed macro block containing a second sub-picture with the first compressed sub-picture without changing the macro block control data, wherein macro block control data is followed by compressed sub-picture blocks to which it relates, wherein the first compressed sub-picture comprises a block motion vector that is user-controllable.
2. A method as claimed in claim 1 , wherein the macro block control data further comprises identification data and the method further comprises the step of selecting the compressed data blocks of the first sub-picture by parsing the compressed video signal to detect identification data corresponding to the first sub-picture.
3. A method as claimed in claim 1 , wherein the compressed video signal comprises a plurality of slices each slice comprising a slice header and a number of consecutive compressed picture blocks.
4. A method as claimed in claim 3 , wherein each slice comprises a number of compressed picture macro blocks corresponding to a width of a picture.
5. A method as claimed in claim 1 , wherein a given sub-picture is predictively coded and wherein all block motion vectors associated with the plurality of picture blocks within the given sub-picture are set to a same value which depends on a displacement of the given sub-picture between a picture and a reference picture.
6. A method as claimed in claim 5 , wherein the sub-picture displacement is user-controllable.
7. A computer program stored on a memory device enabling the carrying out of a method according to claim 1 .
8. An apparatus for manipulating video sub-pictures, comprising:
a mosaic screen generator for receiving an uncompressed video signal from a single source and generating a mosaic screen of horizontally and vertically arranged sub-pictures therefrom;
a compressor for compressing the mosaic screen into compressed macro blocks, each compressed macro block containing only one compressed sub-picture; and
a processor for manipulating a horizontal position of an individual first compressed sub-picture, the processor associating the individual first compressed sub-picture with macro block control data of a compressed macro block containing an individual second compressed sub-picture, and replacing the second compressed sub-picture with the first compressed sub-picture, without changing the macro block control data, wherein macro block control data is followed by compressed sub-picture blocks to which it relates, wherein the individual first compressed sub-picture comprises a block motion vector that is user-controllable.