METHODS AND SYSTEMS OF UPDATING MOTION MODELS FOR OBJECT TRACKERS IN VIDEO ANALYTICS

• US 20180046857A1
• Filed: 12/20/2016
• Published: 02/15/2018
• Est. Priority Date: 08/12/2016
• Status: Active Grant

• Current Assignee: Qualcomm, Inc.
• Original Assignee: Qualcomm, Inc.
• Inventors: Gao, Jinglun, Chen, Ying, Wang, Lei, Bi, Ning
• Granted Patent: US 10,115,005 B2
• Time in Patent Office: Days
• CPC Class Codes: G06T 2207/30232 Surveillance
• CPC Class Codes: G06T 2207/30241 Trajectory
• CPC Class Codes: G06T 7/246 using feature-based methods, e.g. the tracking of corners or segments
• CPC Class Codes: G06V 20/41 Higher-level, semantic clustering, classification or understanding of video scenes, e.g. detection, labelling or Markovian modelling of sport events or news items segmenting video sequences G06V20/49
• CPC Class Codes: G06V 20/52 Surveillance or monitoring of activities, e.g. for recognising suspicious objects recognising microscopic objects G06V20/69
• CPC Class Codes: G06V 40/20 Movements or behaviour, e.g. gesture recognition recognition of facial expressions G06V40/16

Abstract

Techniques and systems are provided for processing video data. For example, techniques and systems are provided for performing context-aware object or blob tracker updates (e.g., by updating a motion model of a blob tracker). In some cases, to perform a context-aware blob tracker update, a blob tracker is associated with a first blob. The first blob includes pixels of at least a portion of one or more foreground objects in one or more video frames. A split of the first blob and a second blob in a current video frame can be detected, and a motion model of the blob tracker is reset in response to detecting the split of the first blob and the second blob. In some cases, a motion model of a blob tracker associated with a merged blob is updated to include a predicted location of the blob tracker in a next video frame. The motion model can be updated by using a previously predicted location of blob tracker as the predicted location of the blob tracker in the next video frame in response to the blob tracker being associated with the merged blob. The previously predicted location of the blob tracker can be determined using a blob location of a blob from a previous video frame.

Claims

1. A method of maintaining one or more motion models for one or more blob trackers, comprising:

associating a first tracker with a first blob, the first blob including pixels of at least a portion of one or more foreground objects in one or more video frames;

detecting a split of the first blob and a second blob in a current video frame; and

resetting a motion model of the first tracker in response to detecting the split of the first blob and the second blob.

2. The method of claim 1, wherein resetting the motion model includes eliminating from the motion model one or more previously determined locations for the first tracker.

3. The method of claim 1, further comprising:

reinitializing the motion model of the first tracker using a location of the first blob in the current video frame.

4. The method of claim 1, wherein the first blob and the second blob are merged into a merged blob in a previous video frame, the previous video frame being obtained earlier in time than the current video frame.

5. The method of claim 4, wherein the first tracker and a second tracker are associated with the merged blob in the previous video frame.

6. The method of claim 1, further comprising:

associating a third tracker with a merged blob in the current video frame, the merged blob including a third blob merged with a fourth blob, and wherein multiple trackers are associated with the merged blob in the current video frame; and

updating a motion model of the third tracker to include a predicted location of the third tracker in a next video frame, wherein the motion model is updated by using a previously predicted location of the third tracker as the predicted location of the third tracker in the next video frame in response to the third tracker being associated with the merged blob, the previously predicted location of the third tracker being determined using a blob location in a previous video frame.

7. The method of claim 6, wherein the third tracker is associated with the third blob in the previous video frame, and wherein a fourth tracker is associated with the fourth blob in the previous video frame.

8. The method of claim 6, wherein the blob location in the previous video frame includes a location of the third blob in the previous video frame before the third blob is merged with the fourth blob.

9. The method of claim 1, wherein the motion model of the first tracker is maintained using a Kalman filter.

10. The method of claim 1, wherein the motion model of the first tracker includes one or more locations of the first tracker, a velocity of the first tracker, and one or more sizes of one or more blobs associated with the first tracker.

11. An apparatus for maintaining one or more motion models for one or more blob trackers comprising:

a memory configured to store video data; and

a processor configured to;

associate a first tracker with a first blob, the first blob including pixels of at least a portion of one or more foreground objects in one or more video frames;

detect a split of the first blob and a second blob in a current video frame; and

reset a motion model of the first tracker in response to detecting the split of the first blob and the second blob.

12. The apparatus of claim 11, wherein resetting the motion model includes eliminating from the motion model one or more previously determined locations for the first tracker.

13. The apparatus of claim 11, wherein the processor is further configured to:

reinitialize the motion model of the first tracker using a location of the first blob in the current video frame.

14. The apparatus of claim 11, wherein the first blob and the second blob are merged into a merged blob in a previous video frame, the previous video frame being obtained earlier in time than the current video frame.

15. The apparatus of claim 14, wherein the first tracker and a second tracker are associated with the merged blob in the previous video frame.

16. The apparatus of claim 11, wherein the processor is further configured to:

associate a third tracker with a merged blob in the current video frame, the merged blob including a third blob merged with a fourth blob, and wherein multiple trackers are associated with the merged blob in the current video frame; and

update a motion model of the third tracker to include a predicted location of the third tracker in a next video frame, wherein the motion model is updated by using a previously predicted location of the third tracker as the predicted location of the third tracker in the next video frame in response to the third tracker being associated with the merged blob, the previously predicted location of the third tracker being determined using a blob location in a previous video frame.

17. The apparatus of claim 16, wherein the third tracker is associated with the third blob in the previous video frame, and wherein a fourth tracker is associated with the fourth blob in the previous video frame.

18. The apparatus of claim 16, wherein the blob location in the previous video frame includes a location of the third blob in the previous video frame before the third blob is merged with the fourth blob.

19. The apparatus of claim 11, wherein the motion model of the first tracker is maintained using a Kalman filter.

20. The apparatus of claim 11, wherein the motion model of the first tracker includes one or more locations of the first tracker, a velocity of the first tracker, and one or more sizes of one or more blobs associated with the first tracker.

21. A computer readable medium having stored thereon instructions that when executed by a processor perform a method, comprising:

associating a first tracker with a first blob, the first blob including pixels of at least a portion of one or more foreground objects in one or more video frames;

detecting a split of the first blob and a second blob in a current video frame; and

resetting a motion model of the first tracker in response to detecting the split of the first blob and the second blob.

22. The computer readable medium of claim 21, wherein resetting the motion model includes eliminating from the motion model one or more previously determined locations for the first tracker.

23. The computer readable medium of claim 21, further comprising:

reinitializing the motion model of the first tracker using a location of the first blob in the current video frame.

24. The computer readable medium of claim 21, wherein the first blob and the second blob are merged into a merged blob in a previous video frame, the previous video frame being obtained earlier in time than the current video frame.

25. The computer readable medium of claim 24, wherein the first tracker and a second tracker are associated with the merged blob in the previous video frame.

26. The computer readable medium of claim 21, further comprising:

associating a third tracker with a merged blob in the current video frame, the merged blob including a third blob merged with a fourth blob, and wherein multiple trackers are associated with the merged blob in the current video frame; and

updating a motion model of the third tracker to include a predicted location of the third tracker in a next video frame, wherein the motion model is updated by using a previously predicted location of the third tracker as the predicted location of the third tracker in the next video frame in response to the third tracker being associated with the merged blob, the previously predicted location of the third tracker being determined using a blob location in a previous video frame.

27. The computer readable medium of claim 26, wherein the third tracker is associated with the third blob in the previous video frame, and wherein a fourth tracker is associated with the fourth blob in the previous video frame.

28. The computer readable medium of claim 26, wherein the blob location in the previous video frame includes a location of the third blob in the previous video frame before the third blob is merged with the fourth blob.

29. The computer readable medium of claim 21, wherein the motion model of the first tracker is maintained using a Kalman filter.

30. The computer readable medium of claim 21, wherein the motion model of the first tracker includes one or more locations of the first tracker, a velocity of the first tracker, and one or more sizes of one or more blobs associated with the first tracker.