Granted Patent B2

US 7,916,796 · App. 11/254,166 · Granted Mar 29, 2011

Abstract

Claims (37)

1. An error detection and correction circuit for a video decoder that reconstructs a lost macroblock of a video frame being decoded, comprising:

a clustering circuit that processes reconstructed frame information of the video frame being decoded, wherein said clustering circuit clusters macroblocks adjacent to the lost macroblock into at least one defined cluster based on at least one attribute of each of said macroblocks adjacent to the lost macroblock;

a classification circuit that assigns the lost macroblock or each of a plurality of sub-blocks of the lost macroblock to a corresponding one of said at least one defined cluster; and

an error concealment circuit that reconstructs attributes of the lost macroblock or of each of said plurality of sub-blocks of the lost macroblock based on selected attributes of said at least one defined cluster to which the lost macroblock or to which each of said plurality of sub-blocks of the lost macroblock is assigned.

2. The error detection and correction circuit of claim 1 , wherein said clustering circuit clusters together each of a plurality of sub-blocks of said macroblocks adjacent to the lost macroblock into said at least one defined cluster.

3. The error detection and correction circuit of claim 1 , wherein said clustering circuit performs clustering using any one or any selected combination of attributes comprising coding parameters, texture statistics, color components, frequency analysis, and image processing operators.

4. The error detection and correction circuit of claim 1 , wherein said classification circuit assigns the lost macroblock in its entirety to said corresponding one of said at least one defined cluster.

5. The error detection and correction circuit of claim 1 , wherein said classification circuit sub-divides the lost macroblock into a plurality of sub-blocks and assigns each sub-block of the lost macroblock to a corresponding one of said at least one defined cluster.

6. The error detection and correction circuit of claim 5 , wherein said classification circuit combines selected ones of said plurality of sub-blocks of the lost macroblock commonly assigned into larger sub-blocks prior to assignment.

7. The error detection and correction circuit of claim 1 , wherein said error concealment circuit reconstructs the lost macroblock by averaging selected attributes of said at least one defined cluster.

8. The error detection and correction circuit of claim 1 , wherein said error concealment circuit reconstructs the lost macroblock by interpolating selected attributes of said at least one defined cluster.

9. The error detection and correction circuit of claim 1 , wherein said error concealment circuit estimates at least one motion vector for the lost macroblock based on a plurality of motion vectors of said at least one defined cluster and receives corresponding motion compensation information.

10. A video decoder using region clustering error concealment to reconstruct a lost macroblock of a video frame, comprising:

a switch having first and second data inputs and an output for providing selected prediction macroblocks;

an intraframe prediction circuit which provides intraframe prediction information to said first input of said switch;

a motion compensation circuit which provides motion compensated prediction information to said second input of said switch;

an adder having a first input coupled to said output of said switch and an output and a second input receiving residual video information, wherein said adder adds selected prediction macroblocks to said residual video information and provides reconstructed frame information at said output of said adder; and

an error detection and correction circuit, having an input receiving said reconstructed frame information, which detects a lost macroblock of said reconstructed frame information, groups macroblocks neighboring said lost macroblock into at least one cluster, assigns said lost macroblock to at least one of said at least one cluster, and reconstructs said lost macroblock based on selected attributes of said at least one cluster to which said lost macroblock is assigned.

11. The video decoder of claim 10 , wherein said error detection and correction circuit comprises:

a clustering circuit that groups said neighboring macroblocks into said at least one cluster;

a classification circuit that assigns said lost macroblock to at least one of said at least one cluster; and

an error concealment circuit that reconstructs said lost macroblock based on selected attributes of said at least one cluster to which said lost macroblock is assigned.

12. The video decoder of claim 11 , wherein said clustering circuit subdivides said neighboring macroblocks into a plurality of neighboring sub-blocks and groups said plurality of neighboring sub-blocks into a plurality of clusters, and wherein said classification circuit subdivides said lost macroblock into a plurality of lost sub-blocks and assigns each said lost sub-block to one of said plurality of clusters.

13. The video decoder of claim 10 , wherein said error detection and correction circuit performs clustering using any one or any selected combination of attributes comprising coding parameters, texture statistics, color components, frequency analysis, and image processing operators.

14. The video decoder of claim 10 , wherein said error detection and correction circuit reconstructs the lost macroblock by averaging each of at least one selected attribute of said at least one cluster.

15. A method of performing error concealment of a lost macroblock of a video frame being decoded by a video decoder, comprising:

region clustering neighboring macroblocks of the video frame being decoded by the video decoder into one or more clusters based on at least one attribute of each of the neighboring macroblocks;

classifying the lost macroblock of the video frame being decoded by the video decoder to the one or more clusters; and

reconstructing the lost macroblock of the video frame being decoded by the video decoder based on selected attributes of the one or more clusters.

16. The method of claim 15 , wherein said region clustering comprises subdividing the neighboring macroblocks into neighboring sub-blocks and grouping the neighboring sub-blocks into the one or more clusters.

17. The method of claim 15 , further comprising:

if more than one cluster is defined, subdividing the lost macroblock into one or more lost sub-blocks;

said classifying comprising assigning each lost sub-block to a corresponding cluster; and

reconstructing each lost sub-block based on selected attributes of a corresponding cluster to which the lost sub-block is assigned.

18. The method of claim 15 , wherein said region clustering comprises clustering based on any one or any selected combination of attributes comprising coding parameters, texture statistics, color components, frequency analysis, and image processing operators.

19. The method of claim 15 , wherein said reconstructing comprises averaging or interpolating attributes.

20. The method of claim 15 , wherein said reconstructing comprises averaging motion vectors and motion compensating.

Assignments (26)