Methods and apparatus for implicit block segmentation in video encoding and decoding

1. An apparatus, comprising:

an encoder for encoding at least a portion of a picture using a block segmentation process, wherein said encoder includes a segmentation device for performing the block segmentation process on a pixel difference of at least two predictors for the portion, each of the at least two predictors corresponding to a respective one of at least two blocks in at least one reference picture and providing a respective predicted pixel value for each of pixels in the block,

wherein said encoder implicitly signals block segmentation information, the at least two predictors, and a weight index for use by a decoder to decode the portion, the block segmentation information indicating which reference picture pixel belongs to which reference picture segment obtained by the block segmentation process, and wherein said weight index indicates weight factors to be used on reference picture segments for calculating a weighted predictor for the portion from a predefined set of weights, and wherein disconnected pixels belonging to a same segment after segmentation are treated as different segments.

2. The apparatus of claim 1 , wherein said encoder signals the at least two predictors for use by the decoder to decode the portion.

3. The apparatus of claim 1 , wherein said encoder signals a weight index for use by the decoder to decode the portion.

4. The apparatus of claim 1 , wherein the block segmentation is performed with respect to at least one block in at least one reference picture.

5. The apparatus of claim 1 , wherein the block segmentation is performed using at least one of edge based segmentation and K-means clustering.

6. The apparatus of claim 1 , wherein after the block segmentation said encoder applies post-processing to the pixel difference of the at least two predictors to remove noisy segments there from and exploit spatial information therein.

7. The apparatus of claim 1 , wherein for each segment, said encoder computes a corresponding weight for each of the at least two predictors, and calculates a final predictor for the portion as a sum of weighted predictors.

8. The apparatus of claim 1 , wherein said encoder signals at least one of motion information and disparity information associated with each of the at least two predictors.

9. The apparatus of claim 1 , wherein the picture is one of a set of pictures corresponding to multi-view video content and having different view points with respect to a same or similar scene, the picture representing a current one of the different view points, wherein at least one of the at least two predictors is an inter-view prediction corresponding to a particular one of the different view points other than the current one of the different view points, and at least another one of the at least two predictors is a temporal prediction corresponding to the current one of the different view points.

10. A method, comprising:

encoding at least a portion of a picture using a block segmentation process, wherein said encoding step includes performing the block segmentation process on a pixel difference of at least two predictors for the portion, each of the at least two predictors corresponding to a respective one of at least two blocks in at least one reference picture and providing a respective predicted pixel value for each of pixels in the block,

wherein said encoding step comprises implicitly signalling block segmentation information, the at least two predictors, and a weight index for use by a decoder to decode the portion, the block segmentation information indicating which reference picture pixel belongs to which reference picture segment obtained by the block segmentation process, and wherein said weight index indicates weight factors to be used on reference picture segments for calculating a weighted predictor for the portion from a predefined set of weights, and wherein disconnected pixels belonging to a same segment after segmentation are treated as different segments.

11. The method of claim 10 , wherein said encoding step comprises signaling the at least two predictors for use by the decoder to decode the portion.

12. The method of claim 10 , wherein said encoding step comprises signaling a weight index for use by the decoder to decode the portion.

13. The method of claim 10 , wherein the block segmentation is performed with respect to at least one block in at least one reference picture.

14. The method of claim 10 , wherein the block segmentation is performed using at least one of edge based segmentation and K-means clustering.

15. The method of claim 10 , further comprising applying post-processing, after the block segmentation, to the pixel difference of the at least two predictors to remove noisy segments there from and exploit spatial information therein.

16. The method of claim 10 , wherein for each segment, said encoding step comprises computing a corresponding weight for each of the at least two predictors, and calculating a final predictor for the portion as a sum of weighted predictors.

17. The method of claim 10 , wherein said encoding step comprises signaling at least one of motion information and disparity information associated with each of the at least two predictors.

18. The method of claim 10 , wherein the picture is one of a set of pictures corresponding to multi-view video content and having different view points with respect to a same or similar scene, the picture representing a current one of the different view points, wherein at least one of the at least two predictors is an inter-view prediction corresponding to a particular one of the different view points other than the current one of the different view points, and at least another one of the at least two predictors is a temporal prediction corresponding to the current one of the different view points.

19. An apparatus, comprising:

a decoder for decoding at least a portion of a picture using a block segmentation process, wherein said decoder includes a segmentation device for performing the block segmentation process on a pixel difference of at least two predictors for the portion, each of the at least two predictors corresponding to a respective one of at least two blocks in at least one reference picture and providing a respective predicted pixel value for each of pixels in the block,

wherein said decoder implicitly determines block segmentation information, the at least two predictors, and a weight index for use by a decoder to decode the portion, the block segmentation information indicating which reference picture pixel belongs to which reference picture segment obtained by the block segmentation process, and wherein said weight index indicates weight factors to be used on reference picture segments for calculating a weighted predictor for the portion from a predefined set of weights, and wherein disconnected pixels belonging to a same segment after segmentation are treated as different segments.

20. The apparatus of claim 19 , wherein the block segmentation information is implicitly determined from at least one syntax element.

21. The apparatus of claim 19 , wherein said decoder receives signaling indicating the at least two predictors for use by the decoder to decode the portion.

22. The apparatus of claim 19 , wherein said decoder receives a weight index for use by the decoder to decode the portion.

23. The apparatus of claim 19 , wherein the block segmentation is performed with respect to at least one block in at least one reference picture.

24. The apparatus of claim 19 , wherein the block segmentation is performed using at least one of edge based segmentation and K-means clustering.

25. The apparatus of claim 19 , wherein after the block segmentation said decoder applies post-processing to the pixel difference of the at least two predictors to remove noisy segments there from and exploit spatial information therein.

26. The apparatus of claim 19 , wherein for each segment, said decoder computes a corresponding weight from weight indices for each of the at least two predictors, and calculates a final predictor for the portion as a sum of weighted predictors.

27. The apparatus of claim 19 , wherein said decoder receives at least one of motion information and disparity information associated with each of the at least two predictors for use in decoding the portion.

28. The apparatus of claim 19 , wherein the picture is one of a set of pictures corresponding to multi-view video content and having different view points with respect to a same or similar scene, the picture representing a current one of the different view points, wherein at least one of the at least two predictors is an inter-view prediction corresponding to a particular one of the different view points other than the current one of the different view points, and at least another one of the at least two predictors is a temporal prediction corresponding to the current one of the different view points.

29. A method, comprising:

decoding at least a portion of a picture using a block segmentation process, wherein said decoding step includes a segmentation device for performing the block segmentation process on a pixel difference of at least two predictors for the portion, each of the at least two predictors corresponding to a respective one of at least two blocks in at least one reference picture and providing a respective predicted pixel value for each of pixels in the block,

wherein said decoding step comprises implicitly determining block segmentation information, the at least two predictors, and a weight index for use by a decoder to decode the portion, the block segmentation information indicating which reference picture pixel belongs to which reference picture segment obtained by the block segmentation process, and wherein said weight index indicates weight factors to be used on reference picture segments for calculating a weighted predictor for the portion from a predefined set of weights, and wherein disconnected pixels belonging to a same segment after segmentation are treated as different segments.

30. The method of claim 29 , wherein the block segmentation information is implicitly determined from at least one syntax element.

31. The method of claim 29 , wherein said decoding step comprises receiving signaling indicating the at least two predictors for use by the decoder to decode the portion.

32. The method of claim 29 , wherein said decoding step comprises receiving a weight index for use by the decoder to decode the portion.

33. The method of claim 29 , wherein the block segmentation is performed with respect to at least one block in at least one reference picture.

34. The method of claim 29 , wherein the block segmentation is performed using at least one of edge based segmentation and K-means clustering.

35. The method of claim 29 , wherein said decoding step comprises applying post-processing, after the block segmentation, to the pixel difference of the at least two predictors to remove noisy segments there from and exploit spatial information therein.

36. The method of claim 29 , wherein for each segment, said decoding step comprises computing a corresponding weight from weight indices for each of the at least two predictors, and calculates a final predictor for the portion as a sum of weighted predictors.

37. The method of claim 29 , wherein said decoding step comprises receiving at least one of motion information and disparity information associated with each of the at least two predictors for use in decoding the portion.

38. The method of claim 29 , wherein the picture is one of a set of pictures corresponding to multi-view video content and having different view points with respect to a same or similar scene, the picture representing a current one of the different view points, wherein at least one of the at least two predictors is an inter-view prediction corresponding to a particular one of the different view points other than the current one of the different view points, and at least another one of the at least two predictors is a temporal prediction corresponding to the current one of the different view points.