Linking tracked objects that undergo temporary occlusion

• US 20040156530A1
• Filed: 08/08/2003
• Published: 08/12/2004
• Est. Priority Date: 02/10/2003
• Status: Active Grant

• Current Assignee: Honeywell International Inc.
• Original Assignee: Honeywell International Inc.
• Inventors: Lin, Yun-Ting, Brodsky, Tomas
• Granted Patent: US 7,394,916 B2
• Time in Patent Office: Days
• US Class Current: 382/103
• CPC Class Codes: G06T 2207/30196 Human being; Person
• CPC Class Codes: G06T 2207/30232 Surveillance
• CPC Class Codes: G06T 7/292 Multi-camera tracking
• CPC Class Codes: G06V 10/24 Aligning, centring, orientation detection or correction of the image

Abstract

A method and system is configured to characterize regions of an environment by the likelihoods of transition of a target from each region to another. The likelihoods of transition between regions is preferably used in combination with conventional object-tracking algorithms to determine the likelihood that a newly-appearing object in a scene corresponds to a recently-disappeared target. The likelihoods of transition may be predefined based on the particular environment, or may be determined based on prior appearances and disappearances in the environment, or a combination of both. The likelihoods of transition may also vary as a function of the time of day, day of the week, and other factors that may affect the likelihoods of transitions between regions in the particular surveillance environment.

Claims

1. A method that facilitates tracking of a target, comprising:

detecting a disappearance of the target from a first region of a scene in a field of view of a camera,

detecting an appearance of an object in a second region of the scene, and

estimating a likelihood of the object corresponding to the target, based on a transition likelihood that is associated with object transitions from the first region to the second region.

2. The method of claim 1, further including:

detecting an appearance of an other object in an other region of the scene, and

estimating a likelihood of the other object corresponding to the target, based on a transition likelihood that is associated with object transitions from the first region to the other region.

3. The method of claim 2, further including

comparing:

a first measure based on the likelihood of the object corresponding to the target to

a second measure based on the likelihood of the other object corresponding to the target,

to affect thereby a determination of whether to link the object or the other object to the target.

4. The method of claim 3, wherein

the determination of whether to link the object or the other object to the target is also based on a likelihood of a further object corresponding to the target.

5. The method of claim 1, further including:

detecting an appearance of an other object in an other region of an other scene in a field of view of an other camera, and

estimating a likelihood of the other object corresponding to the target, based on a transition likelihood that is associated with object transitions from the first region to the other region.

6. The method of claim 1, further including

linking the object as the target for subsequent tracking, based on:

the likelihood of the object corresponding to the target, and

at least one measure of similarity between the object and the target.

7. The method of claim 6, wherein

the at least one measure of similarity is based on at least one of:

colors of the object and the target,

sizes of the object and the target,

shapes of the object and the target,

orientations of the object and the target, and

motions of the object and the target.

8. The method of claim 1, wherein

the disappearance of the target occurs at a first time,

the appearance of the object occurs at a second time, and

the likelihood of the object corresponding to the target is further dependent upon a time duration between the first time and the second time.

9. The method of claim 8, further including

determining the transition likelihood, based on prior disappearances and subsequent appearances in one or more scenes, and further based on time durations between the prior disappearances and subsequent appearances.

10. The method of claim 9, further including

determining the transition likelihood, based on a speed of the object.

11. The method of claim 1, further including

determining the transition likelihood, based on prior disappearances and subsequent appearances in one or more scenes.

12. The method of claim 11, wherein

determining the transition likelihood is further based on a predetermined likelihood of transition between the first region and the second region.

13. The method of claim 12, wherein

determining the transition likelihood is based on a weighted running average that is based on the predetermined likelihood, the prior disappearances and subsequent appearances, and a result of a current disappearance.

14. The method of claim 11, wherein

determining the transition likelihood is based on a ratio of a number of prior transitions from the first region to the second region to a total number of disappearances from the first region.

15. The method of claim 11, wherein

determining the transition likelihood is based on a ratio of a number of prior transitions from the first region to the second region to a total number of appearances at the second region.

16. The method of claim 15, wherein

determining the transition likelihood is further based on a ratio of a number of prior transitions from the first region to the second region to a total number of disappearances from the first region.

17. A method of characterizing a plurality of regions of a scene, to facilitate object-tracking among the plurality of regions, including:

determining a number of transitions of objects from each region to each other region, and

determining a transition likelihood of objects from each region to each other region, based on the number of transitions of objects from each region to each other region.

18. The method of claim 17, further including:

determining at least one of;

a number of arrivals of objects in each region, and

a number of departures of objects in each region; and

wherein

determining the transition likelihood is further based on at least one of the number of arrivals and departures of objects in each region.

19. The method of claim 18, wherein

determining the transition likelihood is further based on a predetermined likelihood of transition between each region and each other region.

20. The method of claim 19, wherein

determining the transition likelihood is based on a weighted running average that is based on the transition likelihood, and a result of a current disappearance.

21. The method of claim 18, wherein

determining the transition likelihood is based on a ratio of the number of transitions from each region to the number of departures from each region.

22. The method of claim 18, wherein

determining the transition likelihood is based on a ratio of the number of transitions to each region to the number of arrivals to each region.

23. The method of claim 22, wherein

determining the transition likelihood is further based on a ratio of a number of transitions from each region to the number of departures from each region.

24. A system that facilitates tracking of objects among a plurality of regions of one or more scenes, comprising:

a video processor that is configured to identify arrivals and departures of objects in each region of the plurality of regions,

a transition predictor, operably coupled to the video processor, that is configured to provide a measure of likelihood that a departure of a first object from a first region of the plurality of regions corresponds to an arrival of a second object at a second region of the plurality of regions,

an object comparer, operably coupled to the video processor, that is configured to provide a measure of similarity between the first object and the second object, and

an object linker, operably coupled to the transition predictor and the object comparer, that is configured to selectively link data associated with the second object to data associated to the first object, based on the measures of likelihood and similarity.

25. The system of claim 24, further including

a display processor, operably coupled to the video processor and the object linker, that is configured to display some or all of the one or more scenes dependent upon the linking of the data associated with the second object to the data associated with the first object.

26. The system of claim 24, wherein

the measure of likelihood is based on prior disappearances and subsequent appearances among the plurality of regions

27. The system of claim 24, wherein

object linker is further configured to selectively link data associated with the second object to data associated to the first object based on measures of likelihood and similarity of other objects to the second object.

28. The system of claim 24, wherein

the measure of similarity is based on at least one of:

colors of the first and second objects,

sizes of the first and second objects,

shapes of the first and second objects,

orientations of the first and second objects, and

motions of the first and second objects.

29. The system of claim 24, wherein

the disappearance of the first object occurs at a first time,

the appearance of the second object occurs at a second time, and

the measure of likelihood is dependent upon a time duration between the first time and the second time.

30. The system of claim 24, wherein

the measure of likelihood is based on prior disappearances and subsequent appearances among the plurality of regions, and further based on time durations between the prior disappearances and subsequent appearances.

31. The system of claim 30, wherein

the measure of likelihood is further based on a speed of the first object.