Method of tracking an object captured by a camera system

• US 8,655,020 B2
• Filed: 07/06/2012
• Issued: 02/18/2014
• Est. Priority Date: 03/03/2008
• Status: Active Grant

• Current Assignee: Motorola Solutions, Inc.
• Original Assignee: VideoIQ, Inc. (Motorola Solutions, Inc.)
• Inventors: Saptharishi, Mahesh, Lisin, Dimitri A.
• Primary Examiner(s): STREGE, JOHN B
• Application Number:
  US13/543,668
• Publication Number:
  US 20120274777A1
• Time in Patent Office: 592 Days
• Field of Search: 382/103, 382/159, 382/224
• US Class Current: 382/103
• CPC Class Codes: G06F 18/22 Matching criteria, e.g. proximity measures
• CPC Class Codes: G06F 2218/12 Classification; Matching
• CPC Class Codes: G06T 2207/10016 Video; Image sequence
• CPC Class Codes: G06T 2207/30232 Surveillance
• CPC Class Codes: G06T 7/246 using feature-based methods, e.g. the tracking of corners or segments
• CPC Class Codes: G06T 7/73 using feature-based methods
• 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: H04N 23/611 where the recognised objects include parts of the human body
• CPC Class Codes: H04N 23/63 by using electronic viewfinders
• CPC Class Codes: H04N 23/631 Graphical user interfaces [GUI] specially adapted for controlling image capture or setting capture parameters
• CPC Class Codes: H04N 23/66 Remote control of cameras or camera parts, e.g. by remote control devices
• CPC Class Codes: H04N 23/661 Transmitting camera control signals through networks, e.g. control via the Internet
• CPC Class Codes: H04N 7/18 Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast
• CPC Class Codes: H04N 7/181 for receiving images from a plurality of remote sources

Abstract

A camera system comprises an image capturing device, object detection module, object tracking module, and match classifier. The object detection module receives image data and detects objects appearing in one or more of the images. The object tracking module temporally associates instances of a first object detected in a first group of the images. The first object has a first signature representing features of the first object. The match classifier matches object instances by analyzing data derived from the first signature of the first object and a second signature of a second object detected in a second image. The second signature represents features of the second object derived from the second image. The match classifier determine whether the second signature matches the first signature. A training process automatically configures the match classifier using a set of possible object features.

Claims

1. A method of tracking an object captured by a camera system, comprising:

receiving image data representing first and second images acquired by the camera system;

detecting a first object captured in the first image and a second object captured in the second image;

producing a first signature of the first object and a second signature of the second object, the first signature representing a first set of features of the first object and the second signature representing a second set of features of the second object;

producing data derived from the first and second signatures; and

analyzing the data with use of a match classifier including a set of match and non-match parameters that populate a feature classification space, and the match classifier including multiple decision steps performed to determine whether the second object is an instance of the first object, each of the multiple decision steps performed by the match classifier analyzing with reference to the set of match and non-match parameters to produce decision results that include accepting the first and second objects as a match or rejecting the first and second objects as a match.

2. The method of claim 1, further comprising determining that the second object is an instance of the first object.

3. The method of claim 1, further comprising producing the second signature in real time.

4. The method of claim 1, further comprising combining portions of the first and second signatures to produce the data.

5. The method of claim 1, wherein the match classifier includes N decision steps configured in a cascade configuration, and each of the N decision steps includes one or more stages for analyzing portions of the data.

8. The method of claim 1, wherein the first signature is a cumulative signature representing a weighted average of single-frame signatures derived from multiple images of the first object, and further comprising, upon determining that the second object is an instance of the first object, updating the first signature of the first object by including the second signature in the cumulative signature.

9. The method of claim 1, wherein the camera system includes first and second image capturing devices connected to each other in a network, further comprising:

capturing with the first image capturing device the first image of the first object; and

capturing with the second image capturing device the second image of the second object.

10. The method of claim 9, wherein the first image capturing device includes the match classifier for analyzing the data, the method further comprising:

transmitting the first signature from the first image capturing device to the second image capturing device.

11. The method of claim 9, wherein a processor remote from the first and second image capturing devices includes the match classifier for analyzing the data, the method further comprising:

transmitting to the processor the first and second signatures from, respectively, the first and second image capturing devices.

12. The method of claim 1, wherein the data correspond to a first subset of features selected from the first set and a second subset of features selected from the second set, and wherein a training process automatically determines the features of the first and second sets that are to be selected for inclusion in the first and second subsets.

13. The method of claim 1, further comprising searching signatures of a set of previously detected objects to identify whether the second object appears in images of the set of objects.

6. A method of tracking an object captured by a camera system, comprising:

receiving image data representing first and second images acquired by the camera system;

detecting a first object captured in the first image and a second object captured in the second image;

producing a first signature of the first object and a second signature of the second object, the first signature representing a first set of features of the first object and the second signature representing a second set of features of the second object;

producing data derived from the first and second signatures, the data corresponding to a first subset of features selected from the first set and a second subset of features selected from the second set, the first subset of features having appearance features including a first edge orientation histogram of the first object, and the second subset of features having appearance features including a second edge orientation histogram of the second object; and

analyzing the data with use of a match classifier that includes N number of multiple decision steps configured in a cascade configuration and performed to determine whether the second object is an instance of the first object, and each of the N decision steps including one or more stages for analyzing portions of the data.

7. The method of claim 6, further comprising mapping a portion of the data derived from the first and second edge orientation histograms to a scalar value representing a measure of similarity.

14. A method of finding an object captured by a camera system, comprising:

receiving image data representing first and second images acquired by the camera system;

detecting a first object captured in the first image and a second object captured in the second image;

producing a first signature of the first object and a second signature of the second object, the first signature representing a first set of features of the first object and the second signature representing a second set of features of the second object;

displaying the second image of the second object on a display;

accepting user input indicative of a selection of the second signature of the second object;

analyzing, with reference to a set of match and non-match parameters that populate a feature classification space of a match classifier, whether the second signature matches the first signature to determine whether the second object is an instance of the first object;

producing a notification indicating a match between the first and second signatures to indicate a match between the first and second objects; and

sending the notification to enable responsive action to the indication of a match.

15. The method of claim 14, further comprising searching signatures of a set of previously detected objects to determine whether the second object matches one of the previously detected objects.

16. The method of claim 15, wherein the set of previously detected objects has been detected by multiple cameras, and the searching includes searching the signatures of the set of previously detected objects detected by the multiple cameras.

17. The method of claim 15, further comprising using selection criteria to filter objects that are members of the set of previously detected objects and that match the second signature of the second object.

18. A method of detecting an object of interest by a camera system, comprising:

receiving image data representing the first and second images acquired by the camera system;

detecting a first object captured in the first image and a second object captured in the second image;

producing a first signature of the first object and a second signature of the second object, the first signature representing a first set of features of the first object and the second signature representing a second set of features of the second object;

displaying the first image of the first object on the display;

accepting user input indicative of a selection of the first object;

analyzing, with reference to a set of match and non-match parameters that populate a feature classification space of a match classifier, whether the second signature matches the first signature to determine whether the second object is an instance of the first object;

producing a notification indicating a match between the first and second signatures to indicate a match between the first and second objects; and

sending the notification to enable responsive action to the indication of a match.

19. The method of claim 18, further comprising monitoring images captured by the camera system in accordance with a rule created in the camera system to watch for objects identified by a signature that matches the first signature.

20. The method of claim 18, wherein the rule created causes the camera system to watch for objects of interest captured in images acquired by multiple cameras.

21. The method of claim 19, wherein the rule specifies criteria for determining when an object is considered an object of interest.

22. The method of claim 18, wherein the selection of the first object includes selection of the first signature of the first object.