Visual sensing for large-scale tracking

• US 7,697,720 B2
• Filed: 09/15/2005
• Issued: 04/13/2010
• Est. Priority Date: 09/18/2004
• Status: Expired due to Fees

• Current Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Original Assignee: Hewlett-Packard Development Company, L.P. (HP Inc.)
• Inventors: Low, Colin Andrew
• Primary Examiner(s): Tucker; Wesley
• Assistant Examiner(s): Bitar; Nancy
• Application Number:
  US11/227,024
• Publication Number:
  US 20070237357A1
• Time in Patent Office: 1,671 Days
• Field of Search: 382/103, 382/165, 382/170, 382/270, 382/107, 382/277, 382/230, 382/304, 382/118, 382/217, 382/173, 348/143, 348/157, 348/169, 348 42- 43, 348/47, 348/51, 348/56, 342/90, 342/96, 345/419, 356/2
• US Class Current: 382/103
• CPC Class Codes: G01S 3/781 Details
• CPC Class Codes: G01S 3/7864 T.V. type tracking systems
• CPC Class Codes: G06T 2207/30232 Surveillance
• CPC Class Codes: G06T 7/292 Multi-camera tracking
• CPC Class Codes: G06V 20/00 Scenes; Scene-specific elements control of digital cameras H04N23/60
• CPC Class Codes: G08B 13/19608 Tracking movement of a target, e.g. by detecting an object predefined as a target, using target direction and or velocity to predict its new position
• CPC Class Codes: G08B 13/19641 Multiple cameras having overlapping views on a single scene

Abstract

One embodiment of a method of tracking a plurality of targets can be broadly summarized by the following steps: capturing a plurality of images of a plurality of targets with a plurality of image capture devices; generating a target observation for each target, said target observation including at least a visual signature of the target and a time value; partitioning target observations according to similarities in their visual signatures; and producing primary tracks from the partitioned target observations, wherein each primary track includes ordered sequences of observation events having similarities in their visual signatures. Other methods and systems are also provided.

Claims

1. A method of tracking a plurality of targets comprises:

capturing a plurality of images of a plurality of targets with a plurality of image capture devices;

generating a target observation for each target, said target observation including at least a visual signature of the target and a time value;

partitioning target observations according to similarities in their visual signatures; and

producing primary tracks from the partitioned target observations, wherein each primary track includes ordered sequences of observation events having similarities in their visual signatures, wherein the primary tracks are decomposed into refined tracks using traffic flow information of events observed in the plurality of images, the traffic information flow being repeatedly revised as additional events are observed.

2. A method of tracking a plurality of targets as claimed in claim 1, in which the primary tracks are decomposed into the refined tracks with a transit graph.

3. A method of tracking a plurality of targets as claimed in claim 2, in which the transit graph represents a historic likelihood of a target moving between two specified locations within a certain period of time.

4. A method of tracking a plurality of targets as claimed in claim 2, in which the traffic flow information comprises a priori information obtained from a network of the image capture devices.

5. A method of tracking a plurality of targets as claimed in claim 4, in which the a priori information is determined by a bootstrap process.

6. A method of tracking a plurality of targets as claimed in claim 4, in which the a priori information is generated using a subset of the target observations.

7. A method of tracking a plurality of targets as claimed in claim 4, in which the primary tracks are decomposed using a subset of the target observations.

8. A method of tracking a plurality of targets as claimed in claim 6, in which said subset has an above average confidence in an accuracy of one or more primary tracks for that subset of target observations.

9. A method of tracking a plurality of targets as claimed in claim 1, in which the visual signature is produced by a colour image segmentation of a target image.

10. A method as claimed in claim 1, in which the visual signature includes information relating to a shape of an identified target.

11. A method of tracking a plurality of targets as claimed in claim 4, in which the a priori information is an average or estimated journey time between two locations or image capture devices, an average or estimated residency time at a location, and/or an indication of whether a location is typically a beginning or an end of the track.

16. A computer readable storage medium storing a computer executable program operable to perform the method of claim 1.

12. Apparatus for tracking a plurality of targets comprises:

a plurality of image capture devices operable to capture a plurality of images of a plurality of targets;

primary computing means operable to

a) generate a target observation for each target;

b) to partition target observations according to similarities in visual signatures of the target observations; and

c) produce primary track estimates from the partitioned target observations, wherein each primary track includes ordered sequences of observation events having similarities in their visual signatures, wherein primary tracks are decomposed into refined tracks using traffic flow information of events observed in the plurality of images, the traffic information flow being repeatedly revised as additional events are observed.

13. Apparatus for tracking a plurality of targets as claimed in claim 12, in which the primary computing means is a computer or a group of linked computers.

14. Apparatus for tracking a plurality of targets as claimed in claim 12, in which the primary computing means incorporates an observation database.

15. Apparatus for tracking a plurality of targets as claimed in claim 12, in which the primary computing means produce target observations based on visual signatures received from the image capture devices.

17. A method of tracking a plurality of targets comprising the steps of:

capturing a plurality of images of a plurality of targets with a plurality of image capture devices;

generating a visual signature and a time value in respect of each instance of observation of a target;

partitioning instances of observation according to similarities in their visual signatures, and

producing primary tracks from the partitioned instances of observations, wherein each primary track includes ordered sequences of observation events having similarities in their visual signatures and primary tracks are decomposed into refined tracks using traffic flow information of events observed in the plurality of images, the traffic information flow being repeatedly revised as additional events are observed.