Video object fragmentation detection and management

• US 8,837,781 B2
• Filed: 10/29/2013
• Issued: 09/16/2014
• Est. Priority Date: 12/23/2008
• Status: Active Grant

• Current Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Original Assignee: Canon Kabushiki Kaisha (Canon Inc.)
• Inventors: Pakulski, Peter Jan, Wedge, Daniel John, Partis, Ashley, See, David Kenji
• Primary Examiner(s): Hung, Yubin
• Application Number:
  US14/065,822
• Publication Number:
  US 20140056477A1
• Time in Patent Office: 322 Days
• Field of Search: 382/103, 348169-172
• US Class Current: 382/103
• CPC Class Codes: G06T 7/248 involving reference images or patches

Abstract

Disclosed herein are a computer-implemented method and a camera system for determining a current spatial representation for a detection in a current frame of an image sequence. The method derives an expected spatial representation (820) for the detection based on at least one previous frame, generates a spatial representation (810) of the detection, and extends the spatial representation (810) to obtain an extended spatial representation (830), based on the expected spatial representation (820). The method determines a similarity measure between the extended spatial representation (830) and the expected spatial representation (820), and then determines the current spatial representation for the detection based on the similarity measure.

Claims

1. A computer-implemented method of tracking an object in an image sequence, said method comprising the steps of:

detecting an object from a current frame in the image sequence;

deriving an expected spatial representation for the object in the current frame based on information of the tracked object obtained from at least one previous frame;

generating a spatial representation of the detected object;

extending the spatial representation of the detected object to obtain an extended spatial representation, based on the expected spatial representation;

calculating a similarity between the extended spatial representation and the expected spatial representation;

determining the detected object as a fragment of the tracked object, if the similarity is above a threshold; and

updating the information of the tracked object using the determined fragment of the tracked object.

2. The method according to claim 1, wherein the spatial representation includes a position of the object.

3. The method according to claim 2, wherein the spatial representation further includes a height and a width of the object.

4. The method according to claim 1, further comprising the step of determining a similarity between the spatial representation of the detected object and the expected spatial representation,

wherein the step of extending the spatial representation of the detected object is performed if the similarity between the spatial representation of the detected object and the expected spatial representation exceeds a threshold.

5. The method according to claim 1, further comprising the step of processing the tracked object.

6. The method according to claim 5, wherein in the processing step, the number of tracked objects are counted.

7. The method according to claim 1, wherein in the extending step, each edge of a bounding box for the spatial representation of the detected object is extended based on a position of a corresponding edge of a bounding box for the expected spatial representation.

8. The method according to claim 1, wherein in the extending step, a width and a height of the spatial representation of the detected object are extended based on a width and a height of the expected spatial representation, respectively.

9. The method according to claim 1, wherein a plurality of objects are tracked and a plurality of objects are detected from the current frame in the detecting step, the method further comprising the step of generating a plurality of association hypotheses, each association hypothesis including an association between one of the tracked objects and at least one of the detected objects.

11. A non-transitory computer readable storage medium storing a program for causing a computer to execute a computer-implemented method defined in claim 1.

10. An apparatus for tracking an object in an image sequence, said apparatus comprising:

a processor;

a memory storing computer instructions when executed by the processor implements;

a detection unit configured to detect an object from a current frame in the image sequence;

a deriving unit configured to derive an expected spatial representation for the object in the current frame based on information of the tracked object obtained from at least one previous frame;

a generation unit configured to generate a spatial representation of the detected object;

an extension unit configured to extend the spatial representation of the detected object to obtain an extended spatial representation, based on the expected spatial representation;

a calculation unit configured to calculate a similarity between the extended spatial representation and the expected spatial representation;

a determination unit configured to determine the detected object as a fragment of the tracked object, if the similarity is above a threshold; and

an updating unit configured to update the information of the tracked object using the determined fragment of the tracked object.