Method and system for object tracking

• US 9,760,791 B2
• Filed: 09/01/2015
• Issued: 09/12/2017
• Est. Priority Date: 09/01/2015
• Status: Active Grant

• Current Assignee: Sony Corporation (Sony Group Corp.)
• Original Assignee: Sony Corporation (Sony Group Corp.)
• Inventors: Rastgar, Houman, Berestov, Alexander
• Primary Examiner(s): Gilliard, Delomia L
• Application Number:
  US14/841,892
• Publication Number:
  US 20170061229A1
• Time in Patent Office: 742 Days
• Field of Search: None
• CPC Class Codes: G06F 18/22 Matching criteria, e.g. proximity measures
• CPC Class Codes: G06T 2207/30196 Human being; Person
• CPC Class Codes: G06T 2207/30232 Surveillance
• CPC Class Codes: G06T 2207/30236 Traffic on road, railway or crossing
• CPC Class Codes: G06T 7/20 Analysis of motion motion estimation for coding, decoding, compressing or decompressing digital video signals H04N19/43, H04N19/51
• CPC Class Codes: G06T 7/248 involving reference images or patches
• CPC Class Codes: G06V 20/40 in video content extracting overlay text G06V20/62; video retrieval G06F16/70; processing of video elementary streams in video servers H04N21/234; processing of video elementary streams in video clients H04N21/44
• CPC Class Codes: G06V 20/52 Surveillance or monitoring of activities, e.g. for recognising suspicious objects recognising microscopic objects G06V20/69

Abstract

Various aspects of a method and system for object tracking are disclosed herein. The method comprises computation of an overlap value between a first bounding box and a second bounding box, associated with an object in an image frame of a video stream. The overlap values are utilized to track the object in the video stream. The first bounding box encompasses at least a partial first portion of an object. The second bounding box encompasses at least a partial second portion of the object. A first score associated with the first bounding box and a second score associated with the second bounding box is computed based on a comparison of the computed overlap value with a pre-defined threshold value. The first bounding box is updated to a new bounding box that at least partially encompasses the object, based on the computed first score and the computed second score.

Claims

1. A method for object tracking, said method comprising:

computing, by an electronic device, an overlap value between a first bounding box and a second bounding box, said first bounding box and said second bounding box associated with an object in an image frame of a video stream,

wherein said first bounding box at least partially encompasses a first portion of said object and said second bounding box at least partially encompasses a second portion of said object, and

wherein said first bounding box and said second bounding box are utilized to track said object in said video stream;

computing, by said electronic device, a first score associated with said first bounding box and a second score associated with said second bounding box based on a comparison of said computed overlap value with a threshold value; and

updating, by said electronic device, said first bounding box to a third bounding box at least partially encompassing said object based on said first score and said second score.

2. The method according to claim 1, further comprising determining an average value of a first set of parameters associated with said first bounding box and a second set of parameters associated with said second bounding box, based on said computed overlap value between said first bounding box and said second bounding box that exceeds said threshold value.

3. The method according to claim 2, wherein said first set of parameters and said second set of parameters comprise at least location information, size information or orientation information.

4. The method according to claim 2, further comprising extracting, by said electronic device, a third portion of said object from said image frame of said video stream corresponding to said determined average value.

5. The method according to claim 4, further comprising updating, by said electronic device, an object model of said object based on content of said third portion.

6. The method according to claim 5, wherein said object model of said object is updated based on a count of instances of said computed overlap value higher than said threshold value by a defined number.

7. The method according to claim 1, further comprising outputting, by said electronic device, said second portion from said image frame of said video stream, based on said computed overlap value between said first bounding box and said second bounding box that is lower than said threshold value and based on said first score that is lower than said second score.

8. The method according to claim 1, wherein said first score associated with said first bounding box is computed based on a motion consistency value between said first bounding box and at least one previous instance of said first bounding box in at least one previous image frame of said video stream.

9. The method according to claim 1, wherein said first score associated with said first bounding box is computed based on a visual similarity value between said object and at least one previous instance of said first bounding box in at least one previous image frame of said video stream.

10. The method according to claim 1, wherein said second score associated with said second bounding box is computed based on a motion consistency value between said second bounding box and at least one of previous instance of said second bounding box in at least one previous image frame of said video stream.

11. The method according to claim 1, wherein said second score associated with said second bounding box is computed based on a visual similarity value between said object and at least one previous instance of said second bounding box in at least one previous image frame of said video stream.

12. The method according to claim 1, wherein said first score and said second score is computed based on said computed overlap value between said first bounding box and said second bounding box that is lower than said threshold value.

13. The method according to claim 12, wherein said first bounding box is updated based on said first score that is lower than said second score.

14. The method according to claim 1, wherein said object is tracked based on at least one attribute of said object, wherein said at least one attribute comprises one of color information of said object or depth information of said object in said image frame.

15. The method according to claim 1, wherein said first bounding box is computed based on a first tracker algorithm.

16. The method according to claim 1, wherein said second bounding box is computed based on a second tracker algorithm.

17. The method according to claim 1, wherein said first portion of said object encompassed by said first bounding box is different from said second portion of said object.

18. The method according to claim 1, wherein said first portion of said object is same as said second portion of said object.

19. A system for object tracking, said system comprising:

one or more circuits in an electronic device, said one or more circuits configured to;

compute an overlap value between a first bounding box and a second bounding box, said first bounding box and said second bounding box associated with an object in an image frame of a video stream,

wherein said first bounding box at least partially encompasses a first portion of said object and said second bounding box at least partially encompasses a second portion of said object, and

wherein said first bounding box and said second bounding box are utilized to track said object in said video stream;

compute a first score associated with said first bounding box and a second score associated with said second bounding box based on a comparison of said computed overlap value with a threshold value; and

update said first bounding box to a third bounding box that at least partially encompasses said object based on said computed first score and said computed second score.

20. A non-transitory computer-readable storage medium having stored thereon, computer-executable instructions for causing a computer to execute operations, said operations comprising:

computing, by an electronic device, an overlap value between a first bounding box and a second bounding box, said first bounding box and said second bounding box associated with an object in an image frame of a video stream,

wherein said first bounding box at least partially encompasses a first portion of said object and said second bounding box at least partially encompasses a second portion of said object, and

wherein said first bounding box and said second bounding box are utilized to track said object in said video stream;

computing, by said electronic device, a first score associated with said first bounding box and a second score associated with said second bounding box based on a comparison of said computed overlap value with a threshold value; and

updating, by said electronic device, said first bounding box to a third bounding box at least partially encompassing said object based on said computed first score and said computed second score.