Signaling valid entry points in a video stream

• US 7,839,930 B2
• Filed: 06/30/2004
• Issued: 11/23/2010
• Est. Priority Date: 11/13/2003
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Srinivasan, Sridhar, Crinon, Regis J., Regunathan, Shankar, Holcomb, Thomas W., Onders, Timothy E.
• Primary Examiner(s): Dastouri; Mehrdad
• Assistant Examiner(s): Werner; David N
• Application Number:
  US10/882,739
• Publication Number:
  US 20050105883A1
• Time in Patent Office: 2,337 Days
• Field of Search: 386/68, 386/69, 386/52, 375/240.26, 375/240.29
• US Class Current: 375/240.13
• CPC Class Codes: H04N 21/23424 involving splicing one content stream with another content stream, e.g. for inserting or substituting an advertisement
• CPC Class Codes: H04N 21/44016 involving splicing one content stream with another content stream, e.g. for substituting a video clip
• CPC Class Codes: H04N 21/8455 involving pointers to the content, e.g. pointers to the I-frames of the video stream

Abstract

Techniques and tools for coding/decoding of digital video, and in particular, for determining, signaling and detecting entry points in video streams are described. Techniques and tools described herein are used to embed entry point indicator information in the bitstream that receivers, editing systems, insertion systems, and other systems can use to detect valid entry points in compressed video.

Claims

1. In a computing device that implements a video encoder, a method comprising:

with the computing device that implements the video encoder, determining one or more entry points for a sequence comprising plural video frames based at least in part on an entry point policy; and

with the computing device that implements the video encoder, inserting in a bitstream an entry point indicator that indicates a first frame of the plural frames is one of the one or more entry points, wherein the first frame is a field interlaced P/I-frame comprising a P-field followed by an I-field, the P-field and the I-field being arranged in the bitstream for decoding as individual fields, wherein a second frame immediately follows the field interlaced P/I-frame in presentation time, wherein the second frame is a predicted frame, wherein according to the entry point policy the video encoder makes coding decisions for the second frame to remove any motion compensation dependencies prior to the I-field of the field interlaced P/I-frame, and wherein the entry point policy comprises conditions that include;

after the field interlaced P/I-frame, at least one field interlaced P/P-frame is present in the bitstream before a progressive P-frame or frame interlaced P-frame;

after the field interlaced P/I-frame, a next P-field in the bit stream references the I-field of the field interlaced P/I-frame but not the P-field of the field interlaced P/I-frame;

after the field interlaced P/I-frame, a progressive B-frame or frame interlaced B-frame in the bitstream that has a later presentation time than the field interlaced P/I-frame is not encoded depending on the field interlaced P/I-frame; and

after the field interlaced P/I-frame, a next B-field in the bitstream that has a later presentation time than the field interlaced P/I-frame is not encoded depending on the P-field of the field interlaced P/I-frame.

2. The method of claim 1 wherein the one or more entry points comprises plural entry points positioned at different intervals.

3. The method of claim 1 wherein the entry point indicator comprises an entry point header.

4. The method of claim 3 wherein the entry point header includes buffer fullness information.

5. The method of claim 1 wherein the entry point indicator comprises an entry point start code.

6. The method of claim 5 wherein the entry point start code is an abbreviated start code sufficient to distinguish the entry point start code from other start codes in the bitstream.

7. The method of claim 1 wherein the bitstream has plural syntax levels, and wherein the entry point indicator is included at a syntax level below sequence level.

8. The method of claim 7 wherein the syntax level below sequence level is entry point level.

9. The method of claim 1 wherein the entry point indicator immediately follows a sequence header in the bitstream.

10. The method of claim 1 wherein the entry point indicator indicates a commercial insertion point to facilitate video splicing.

11. The method of claim 1 wherein, due to designation of the field interlaced P/I-frame as an entry point, a field of the second frame that might otherwise use the P-field of the field interlaced P/I-frame as a reference field in motion compensation is constrained to not use the P-field of the field interlaced P/I-frame as a reference field in motion compensation.

12. The method of claim 11 wherein, due to the designation of the field interlaced P/I-frame as an entry point, the second frame is constrained to be a P/P-frame instead of a progressive P-frame or interlaced P-frame referencing the entire field interlaced P/I-frame.

13. The method of claim 1 wherein the coding decisions include frame coding type of the second frame, picture type for the second frame, and reference picture selection for motion compensation for the second frame,

14. The method of claim 1 wherein a B-field in the sequence precedes the field interlaced P/I-frame in presentation time but follows the field interlaced P/I frame in coded order and uses the P-field of the field interlaced P/I-frame as a reference field.

15. The method of claim 1 wherein a B-frame in the sequence precedes the field interlaced P/I-frame in presentation but follows the field interlaced P/I frame in coded order and uses the field interlaced P/I-frame as a reference frame.

16. In a computing device that implements a video decoder, a method comprising:

receiving, at the computing device that implements the video decoder, a bitstream for a sequence comprising plural video frames, each of the plural video frames having a frame coding type, wherein the frame coding type for each of the plural video frames is one of progressive, field interlaced, and frame interlaced;

with the computing device that implements the video decoder, detecting an entry point indicator in the bitstream, wherein the entry point indicator marks a first frame of the plural frames as an entry point for decoding and is positioned in the bitstream according to an entry point policy; and

with the computing device that implements the video decoder, decoding at least a portion of the bitstream based at least in part on the entry point indicator;

wherein the entry point policy comprises one or more conditions for entry point indicator positioning, wherein the first frame is a field interlaced P/I-frame comprising a P-field followed by an I-field, the P-field and the I-field being arranged in the bitstream for decoding as individual fields, wherein a second frame immediately follows the field interlaced P/I-frame in presentation time, and wherein the frame coding type of the second frame depends in part on the field interlaced P/I frame being an entry point; and

wherein the one or more conditions of the entry point policy comprise;

after the field interlaced P/I-frame, at least one field interlaced P/P-frame is present in the bitstream before a progressive P-frame or frame interlaced P-frame;

after the field interlaced P/I-frame, a next P-field in the bit stream references the I-field of the field interlaced P/I-frame but not the P-field of the field interlaced P/I-frame;

after the field interlaced P/I-frame a progressive B-frame or, frame interlaced B-frame in the bitstream that has a later presentation time than the field interlaced P/I-frame is not encoded depending on the field interlaced P/I-frame; and

after the field interlaced P/I-frame, a next B-field in the bitstream that has a later presentation time than the field interlaced P/I-frame is not encoded depending on the P-field of the field interlaced P/I-frame.

17. The method of claim 16 wherein the bitstream comprises plural syntax levels, and wherein the entry point indicator is at a syntax level below sequence level.

18. The method of claim 17 wherein the syntax level below sequence level is entry point level.

19. The method of claim 16 wherein the entry point indicator comprises an entry point start code and an entry point header.

20. The method of claim 16, the one or more conditions further comprising:

a frame interlaced P-frame following the field interlaced P/I-frame is not predicted from the field interlaced P/I-frame.

21. The method of claim 16 wherein a frame following the field interlaced P/I-frame is a frame selected from the set comprising:

frame interlaced I-frame, field interlaced I/P-frame, field interlaced I/I-frame, progressive I-frame and field interlaced P/I-frame.

22. The method of claim 16 further comprising, with the computing device that implements the video decoder, producing a summarized version of the sequence based at least in part on the entry point.

23. The method of claim 22 wherein the producing a summarized version is performed via a video editing application.

24. The method of claim 16 wherein the decoding is performed in a trick mode.

25. The method of claim 24 wherein the trick mode is fast rewind mode.

26. The method of claim 24 wherein the trick mode is fast forward mode.

27. The method of claim 16 wherein, due to designation of the field interlaced P/I-frame as an entry point, a field of the second frame that might otherwise use the P-field of the field interlaced P/I-frame as a reference field in motion compensation is constrained to not use the P-field of the field interlaced P/I-frame as a reference field in motion compensation.

28. The method of claim 27 wherein, due to the designation of the field interlaced P/I-frame as an entry point, the second frame is constrained to be a P/P-frame instead of a progressive P-frame or interlaced P-frame referencing the entire field interlaced P/I-frame.

29. The method of claim 16 wherein the second frame is a predicted frame, and wherein one or more picture types for the second frame and reference picture selection for motion compensation for the second frame also depend in part on the field interlaced P/I frame being an entry point.

30. The method of claim 16 wherein a B-field in the sequence precedes the field interlaced P/I-frame in presentation but follows the field interlaced P/I frame in coded order and uses the P-field of the field interlaced P/I-frame as a reference field.

31. The method of claim 16 wherein a B-frame in the sequence precedes the field interlaced P/I-frame in presentation but follows the field interlaced P/I frame in coded order and uses the field interlaced P/I-frame as a reference frame.

32. In a computing device that implements a video decoder, a method comprising:

receiving, at the computing device that implements the video decoder, a bitstream for a sequence comprising plural video frames;

with the computing device that implements the video decoder, detecting an entry point indicator in the bitstream, wherein a first frame immediately following the entry point indicator is decodable without reference to any frames preceding the entry point indicator in the bitstream, wherein the first frame is a field interlaced P/I-frame comprising a P-field followed by an I-field, the P-field and the I-field being arranged in the bitstream for decoding as individual fields; and

with the computing device that implements the video decoder, decoding at least a portion of the bitstream based at least in part on the entry point indicator, wherein, due to designation of the field interlaced P/I-frame as an entry point, a field following the field interlaced P/I-frame that might otherwise use the P-field of the field interlaced P/I-frame as a reference field in motion compensation is constrained to not use the P-field of the field interlaced P/I-frame as a reference field in motion compensation, and wherein the bitstream complies with the following conditions;

after the field interlaced P/I-frame. at least one field interlaced P/P-frame is present in the bitstream before a progressive P-frame or frame interlaced P-frame;

after the field interlaced P/I-frame, a next P-field in the bit stream references the I-field of the field interlaced P/I-frame but not the P-field of the field interlaced P/I-frame;

after the field interlaced P/I-frame, a progressive B-frame or frame interlaced B-frame in the bitstream that has a later presentation time than the field interlaced P/I-frame is not encoded depending on the field interlaced P/I-frame; and

after the field interlaced P/I-frame, a next B-field in the bitstream that has a later presentation time than the field interlaced P/I-frame is not encoded depending on the P-field of the field interlaced P/I-frame.

33. The method of claim 32 wherein the detecting the entry point indicator comprises detecting an entry point start code.

34. The method of claim 32 wherein the entry point indicator comprises an entry point start code and an entry point header.

35. The method of claim 34 wherein the entry point header comprises buffer fullness information.

36. In a computing device that implements a video decoder, a method comprising:

receiving, at the computing device that implements the video decoder, a bitstream for a sequence comprising plural video frames; and

with the computing device that implements the video decoder, decoding starting from an entry point signaled by an entry point indicator in the bitstream, wherein the entry point is a field interlaced P/I-frame comprising a P-field followed by an I-field within the plural video frames, wherein after he entry point indicator the bitstream includes the following elements for the entry point;

frame header, first field header, first field data, second field header, and second field data, and wherein the first field header and the first field data are for the P-field and the second field header and the second field data are for the I-field, the frame header being organized according to a different bitstream syntax than the first and second field headers.

37. The method of claim 36 wherein the bitstream further includes a frame start code immediately before the frame header and a field start code immediately before the second field header.

38. The method of claim 36 wherein the entry point indicator includes an entry point start code and an entry point header.