Method and device for temporal segmentation of a video sequence

• US 20030132955A1
• Filed: 01/10/2003
• Published: 07/17/2003
• Est. Priority Date: 01/16/2002
• Status: Active Grant

• Current Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Original Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Inventors: Le Floch, Herve
• Granted Patent: US 7,228,502 B2
• Time in Patent Office: Days
• US Class Current: 345/723
• CPC Class Codes: G11B 27/28 by using information signals recorded by the same method as the main recording G11B27/22 takes precedence
• CPC Class Codes: H04N 19/467 characterised by the embedded information being invisible, e.g. watermarking
• CPC Class Codes: H04N 19/587 involving temporal sub-sampling or interpolation, e.g. decimation or subsequent interpolation of pictures in a video sequence

Abstract

According to this method of temporal segmentation of a video sequence (200) containing one or more markers, the analysis of the content of these markers makes it possible to segment the video sequence into a plurality of video sub-sequences.

Claims

1. A method of temporal segmentation of a video sequence (200) containing one or more markers, characterized in that the analysis of the content of said markers makes it possible to segment the video sequence into a plurality of video sub-sequences.

2. A method according to claim 1, characterized in that it comprises steps of:

dividing (201) the video sequence (201) into a plurality of groups of pictures;

successively selecting (202) each group of said plurality of groups of pictures;

extracting (203) a binary message from each selected group of pictures;

successively analyzing (204) the validity of each binary message extracted, so as to determine whether said message is a marker or a sequence of natural noise; and

grouping together said groups of pictures according to the content of their message.

3. A method according to claim 1 or 2, characterized in that, at the extraction step (203), a local marker (402) and a global marker (404) are predicted and a first message (408) is extracted from the global marker (404) and a second message (407) is extracted from the local marker (402).

4. A method according to claim 1, 2 or 3, characterized in that, at the extraction step (203), the size of said message is calculated and the message is then extracted on the basis of said size.

5. A method according to any one of the preceding claims, characterized in that in addition a video synchronization step is carried out.

6. A method according to any one of the preceding claims, characterized in that, at the analysis step (204), calculation (410, 411;

506;

609) is made of a measure of confidence in the hypothesis that the binary message comes from a group of pictures containing a marker.

7. A method according to any one of the preceding claims, characterized in that, at the analysis step (204), it is tested (510;

610) whether the message extracted from the current group of pictures is identical to the message extracted from the preceding group of pictures.

8. A method according to the preceding claim, characterized in that, in order to test whether the message extracted from the current group of pictures is identical to the message extracted from the preceding group of pictures, a calculation (1105) is made of the percentage (P) of bits that are different between the message extracted from the current group of pictures and the message extracted from the preceding group of pictures and it is decided (1107) that these messages are identical if said percentage (P) is less than a predetermined first value.

9. A method according to claim 7, characterized in that, in order to test whether the message extracted from the current group of pictures is identical to the message extracted from the preceding group of pictures, a calculation (1105) is made of the percentage (P) of bits that are different between the message extracted from the current group of pictures and the message extracted from the preceding group of pictures and it is decided (1107) that these messages are identical if said percentage (P) is less than a predetermined first value and if no temporal de-synchronization is noted or if said percentage (P) is greater than a second predetermined value and if temporal de-synchronization is noted.

10. A method according to any one of the preceding claims, characterized in that, at the analysis step (204), the statistic of the errors between a reference message and local messages is studied (802, 803, 804), and the status and the message of groups of pictures initially detected as not comprising any marker is modified (805, 806, 807, 808, 809), if the percentage of errors between the local message considered and the reference message is sufficiently close to said statistic, within the meaning of a predetermined criterion.

11. A method according to the preceding claim, characterized in that, at the analysis step (204), the average number of erroneous bits between the reference message and the local messages is calculated.

12. A method according to claim 10 or 11, characterized in that the reference message is calculated from a global group of pictures preceding a local group of pictures for which a change of message has been detected.

13. A method according to claim 11 or 12, characterized in that it further comprises steps of:

comparing (808) the average number of erroneous bits with a binary error rate calculated beforehand, and

if the average number of erroneous bits is less than the weighted binary error rate by a predetermined value, the status of the corresponding group of pictures is modified (809) so that, if it was considered as not comprising any marker, it will be considered after modification as comprising a marker.

14. A method according to any one of the preceding claims, characterized in that, at the analysis step (204), the statuses and the messages linked to local groups of pictures are improved by a feedback loop.

15. A method according to any one of the preceding claims, characterized in that each group of pictures is constituted by 24 consecutive pictures.

31. A digital signal processing apparatus, characterized in that it comprises means adapted to implement a method of analysis according to any one of claims 1 to 15.

16. A device for temporal segmentation of a video sequence (200) containing one or more markers, characterized in that it comprises means for analyzing the content of said markers making it possible to segment the video sequence into a plurality of video sub-sequences.

17. A device according to claim 16, characterized in that it comprises:

means for dividing the video sequence (200) into a plurality of groups of pictures;

means for successively selecting each group of said plurality of groups of pictures;

means for extracting a binary message from each selected group of pictures;

means for successively analyzing the validity of each binary message extracted, so as to determine whether said message is a marker or a sequence of natural noise; and

means for grouping together said groups of pictures according to the content of their message.

18. A device according to claim 16 or 17, characterized in that said extraction means are adapted to predict a local marker (402) and a global marker (404) and to extract a first message (408) from the global marker (404) and a second message (407) from the local marker (402).

19. A device according to claim 16, 17 or 18, characterized in that said extraction means are adapted to calculate the size of said message and then to extract the message on the basis of said size.

20. A device according to any one of claims 16 to 19, characterized in that it further comprises video synchronization means.

21. A device according to any one of claims 16 to 20, characterized in that said analysis means are adapted to calculate a measure of confidence in the hypothesis that the binary message comes from a group of pictures containing a marker.

22. A device according to any one of claims 16 to 21, characterized in that said analysis means comprise means for testing whether the message extracted from the current group of pictures is identical to the message extracted from the preceding group of pictures.

23. A device according to the preceding claim, characterized in that said testing means are adapted to calculate the percentage (P) of bits that are different between the message extracted from the current group of pictures and the message extracted from the preceding group of pictures and to decide that these messages are identical if said percentage (P) is less than a first predetermined value.

24. A device according to claim 22, characterized in that said testing means are adapted to calculate the percentage (P) of bits that are different between the message extracted from the current group of pictures and the message extracted from the preceding group of pictures and to decide that these messages are identical if said percentage (P) is less than a first predetermined value and if no temporal de-synchronization is noted or if said percentage (P) is greater than a second predetermined value and if temporal de-synchronization is noted.

25. A device according to any one of claims 16 to 24, characterized in that said analysis means comprise means for studying the statistic of the errors between a reference message and local messages, and means for modifying the status and the message of groups of pictures initially detected as not comprising any marker, if the percentage of errors between the local message considered and the reference message is sufficiently close to said statistic, within the meaning of a predetermined criterion.

26. A device according to the preceding claim, characterized in that said analysis means comprise means for calculating the average number of erroneous bits between the reference message and the local messages.

27. A device according to claim 25 or 26, characterized in that the reference message is calculated from a global group of pictures preceding a local group of pictures for which a change of message has been detected.

28. A device according to claim 26 or 27, characterized in that it further comprises:

means for comparing the average number of erroneous bits with a binary error rate calculated beforehand, and

means for modifying, if the average number of erroneous bits is less than the weighted binary error rate by a predetermined value, the status of the corresponding group of pictures, so that, if it was considered as not comprising any marker, it will be considered after modification as comprising a marker.

29. A device according to any one of claims 16 to 28, characterized in that the analysis means implement a feedback loop adapted to improve the statuses and the messages linked to local groups of pictures.

30. A device according to any one of the claims 16 to 29, characterized in that each group of pictures is constituted by 24 consecutive pictures.

32. A digital signal processing apparatus, characterized in that it comprises an analysis device according to any one of claims 16 to 30.