METHOD AND SYSTEM FOR ENCODING A 3D VIDEO SIGNAL, ENCLOSED 3D VIDEO SIGNAL, METHOD AND SYSTEM FOR DECODER FOR A 3D VIDEO SIGNAL

• US 20100195716A1
• Filed: 06/19/2008
• Published: 08/05/2010
• Est. Priority Date: 06/26/2007
• Status: Active Grant

• Current Assignee: Koninklijke Philips Electronics N.V. (Koninklijke Philips N.V.)
• Original Assignee: Koninklijke Philips Electronics N.V. (Koninklijke Philips N.V.)
• Inventors: Varekamp, Christiaan, Klein Gunnewiek, Reinier Bernardus Maria, Bruls, Wilhelmus Hendrikus Alfonsus
• Granted Patent: US 8,345,751 B2
• Time in Patent Office: Days
• US Class Current: 375/240.08
• CPC Class Codes: H04N 19/597 specially adapted for multi-view video sequence encoding

Abstract

In a method for encoding and an encoder for a 3D video signal, centre view frames, a depth map for centre view frames and an occlusion data frame are encoded. On the basis of the depth map for the centre view frame a distinction is made between functional and non-functional data in an occlusion data frame. This allows a strong reduction in bits needed for the encoded occlusion data frame. In the decoder a combined data stream is made of functional data in the encoded occlusion data frames and the centre view frames. Preferably the centre view frames are used as reference frames in encoding the occlusion data frames.

Claims

1. Method for encoding 3D video signals wherein a 3D video signal is encoded, the 3D video signal comprising a centre view video frame (61), a depth map for the centre view video frame (62) and an occlusion data frame for the center view video frame (63), wherein, under control of data in the depth map for the centre view, for the occlusion data frame data an indication means is generated for distinguishing functional data in the occlusion data frame from nonfunctional data prior to encoding the occlusion data frame (63).

2. Method for encoding as claimed in claim 1, wherein the occlusion data frame (63) is encoded after the depth map (63) is encoded.

3. Method for encoding as claimed in claim 1, where also an occlusion depth map is encoded.

4. Method for encoding as claimed in claim 3, where occlusion depth frame data is used for generation of an indication means for distinguishing functional data in the occlusion data frame from nonfunctional data.

5. Method for encoding as claimed in claim 1, wherein under control of data in the depth map a fixed value or a value within a fixed range is generated in the occlusion data frame as an indication means to distinguish non-functional data (occl=“

0”

).

6. Method for encoding as claimed in claim 5 wherein the fixed value or value within a fixed range is reserved for non-functional data, and any functional data of the fixed value or within the fixed value range is replaced with a value other than the reserved value or outside the reserved range.

7. Method for encoding as claimed in claim 1, wherein in the control a depth discontinuity threshold is applied, wherein a depth discontinuity in the depth map being smaller than a threshold value (TH) is a determining factor.

8. Method for encoding as claimed in claim 7, wherein the threshold value (TH) is calculated from the data in the depth map.

9. Method for encoding as claimed in claim 5, wherein the threshold value (TH) and/or the fixed value or fixed range is incorporated in the encoded video signal.

10. Method for encoding as claimed in claim 9 wherein the threshold value (TH) and/or the fixed value or fixed range is incorporated as metadata in the video signal.

11. Method for encoding as claimed in claim 1 wherein multiple occlusion parts are combined into one occlusion data frame and indication means are generated for distinguishing the different occlusion parts from each other.

12. Method for encoding as claimed in claim 11 wherein transitions occurrences from functional (occl≠

“

0”

) to non-functional (occl=“

0”

) and/or from non-functional (occl=“

0”

) to functional (occl≠

“

0”

) in the occlusion signal for a least one pixel is used as an indicator for distinguishing the different occlusion parts within the occlusion data frame.

13. Method for encoding as claimed in claim 1 wherein a further determining factor in the control is the maximum viewing angle.

14. Method for encoding as claimed in claim 1, wherein a further determining factor in the control is formed by the bit rate of the coded occlusion data frame.

15. Method as claimed in claim 1 wherein center view data frames are used as reference frames in encoding of occlusion data frames.

16. Method as claimed in claim 15 wherein time shifted centre view data are used in encoding of occlusion data frames and the time shifts are calculated from the image data.

17. Method as claimed in claim 16, wherein the calculated time shifts are incorporated into the video signal.

18. Method as claimed in claim 17, wherein the calculated time shifts are incorporated as metadata into the video signal.

45. Computer program comprising program code means for performing a method for encoding as claimed in claim 1 when said program is run on a computer.

47. Computer program product comprising program code means stored on a computer readable medium for performing an encoding method as claimed in claim 1.

19. Encoder for encoding a 3-D video signal, the encoded 3D video signal comprising a centre view video frame (61), a depth map (62) for the centre view video frame and an occlusion data frame (63) for the center view video frame, wherein the encoder comprises inputs for to be encoded centre view data frames, depth maps for a centre view data frame, and occlusion data frames, and the encoder comprises a controller (65), which controller, on the basis of data for depth map, controls a switch (66), said switch generating for the occlusion data frame data either non-functional data (occl=“

0”

) or functional data.

20. Encoder as claimed in claim 19, wherein the encoder comprises an aligner and the alignment of the functional and nonfunctional data is blockgrid aligned.

21. Encoder as claimed in claim 20, wherein the aligner is arranged for aligning the blockgrid at 4×

4, 8×

8, 16×

16 and/or any other integer combination.

22. Encoder as claimed in claim 19, wherein the encoder comprises an encoder (Encoder CV(z)) for encoding the depth map for the centre view frames and an internal decoder (decoder CV(Z)) for decoding the encoding depth maps, wherein the depth maps decoded by the internal decoder are used in the controller (65).

23. Encoder as claimed in claim 19, wherein the encoder comprises a calculator for calculation a threshold value (TH) for the depth discontinuity in the depth map, wherein the control of the switch is dependent on a measured depth discontinuity versus threshold value comparison.

24. Encoder as claimed in claim 23 wherein the encoder comprises an output for incorporating the calculated threshold (TH) into the encoded data stream.

25. Encoder as claimed in claim 19 wherein the encoder comprises an encoder for encoding centre view frames and a decoder for decoding encoded centre view frames, an encoder for encoding the occlusion data frames and an input at the encoder for encoding the occlusion data frames for inserting decoded centre view frames as reference frames.

26. Encoder as claimed in claim 23, wherein shift parameters (τ

2, a) are calculated and incorporated into the video signal.

27. Method for decoding an encoded video signal wherein a 3D video signal is decoded, the 3D video signal comprising an encoded centre view video frame, a depth map for the centre view video frame and an occlusion data frame for the center view video frame, the encoded video signal comprising indications to distinguish non-functional occlusion data from functional occlusion data wherein in the decoding a combined data stream (78, 88) is produced, comprising functional occlusion data and centre view data.

28. Method for decoding an encoded video signal as claimed in claim 27 wherein in the decoding of the occlusion data frames centre view data frames are used as reference frames.

29. Method for decoding an encoded video signal as claimed in claim 27 wherein the combined signal (88) is warped after generation.

30. Method for decoding an encoded signal as claimed in claim 27, wherein warping is performed prior to generation of the combined signal (78).

31. Method for decoding an encoded signal as claimed in claim 27 wherein in the decoding a combined depth data stream (88z) is produced, comprising functional depth occlusion data and depth centre view data.

46. Computer program comprising program code means for performing a method for decoding as claimed in claim 27 when said program is run on a computer.

48. Computer program product comprising program code means stored on a computer readable medium for performing a decoding method as claimed in claim 27.

32. Decoder for decoding an encoded video signal wherein a 3D video signal is decoded, the 3D video signal comprising an encoded centre view video frame, a depth map for the centre view video frame and an occlusion data frame for the center view video frame, the encoded video signal comprising indications to distinguish non-functional occlusion data from functional occlusion data wherein the decoder comprises a combiner (76, 86) for producing a combined data stream (78, 88), said combined signal comprising functional occlusion data and centre view data.

33. Decoder for decoding an encoded video signal as claimed in claim 32 wherein the decoder for the occlusion data frames is arranged for using in the decoding of the occlusion data frames centre view data frames used as reference frames.

35. Decoder for decoding an encoded signal as claimed in claim 32 wherein the decoder comprises a warper (warp A) for performing warping prior to generation of the combined signal (78).

36. Decoder for decoding an encoded signal as claimed in claim 32 wherein the decoder comprises a combiner for producing a combined depth data stream (88z), comprising functional depth occlusion data and depth centre view data.

37. Decoder for decoding an encoded signal as claimed in claim 32 wherein the encoded signal comprises multiple occlusion parts combined into one occlusion data frame and the encoded signal comprises indicators for distinguishing the different occlusion parts from each other, the decoder having means for identifying the indicators and using the indicators identifying the various occlusion parts.

38. Decoder for decoding an encoded signal as claimed in claim 37 wherein the indicators are formed by transitions occurrences from functional (occl≠

“

0”

) to non-functional (occl=“

0”

) and/or from non-functional (occl=“

0”

) to functional (occl#“

0”

) in the occlusion signal.

34. Decoder for decoding an encoded video signal as claimed in 32 wherein the decoder comprises a warper (warp 1) for warping the combined signal (88).

39. 3D video signal comprising an encoded centre view video frame, a depth map for the centre view video frame and an occlusion data frame for the center view video frame, wherein the 3D video signal comprises an indication means for distinguishing non-functional occlusion data from functional occlusion data.

40. 3D video signal as claimed in claim 39 wherein the nonfunctional data have a fixed value or a value within a fixed range and the functional values have a value other than the fixed value or a value outside the fixed range, said value acting as indication means.

41. 3D video signal as claimed in claim 39 wherein the indication means comprises a threshold value for the depth discontinuity.

42. 3D video signal as claimed in claim 39 wherein the 3D video signal comprises an encoded depth map for the occlusion data frame, the encoded depth map comprising the indication means for distinguishing non-functional occlusion data from functional occlusion data.

43. 3D video signal as claimed in claim 39 wherein the video signal comprises, as metadata, a range of values or a value for the occlusion data, said value or range indicating whether occlusion data is functional or nonfunctional.

44. 3D video signal as claimed in claim 39 wherein the video signal comprises control information for controlling the manner in which, during decoding centre view frames are used as reference frames for decoding occlusion data frames.