System, method and article of manufacture for tracking a head of a camera-generated image of a person

• US 20050185054A1
• Filed: 04/21/2005
• Published: 08/25/2005
• Est. Priority Date: 07/30/1999
• Status: Active Grant

• Current Assignee: DigiMedia Tech, LLC (IP Investments Group LLC)
• Original Assignee: Planet Electric Incorporated
• Inventors: Edwards, Jeffrey L., Nguyen, Katerina H.
• Granted Patent: US 7,715,476 B2
• Time in Patent Office: Days
• US Class Current: 348/169
• CPC Class Codes: G06T 7/246 using feature-based methods, e.g. the tracking of corners or segments

Abstract

A system, method and article of manufacture are provided for tracking a head portion of a person image in video images. Upon receiving video images, a first head tracking operation is executed for generating a first confidence value. Such first confidence value is representative of a confidence that a head portion of a person image in the video images is correctly located. Also executed is a second head tracking operation for generating a second confidence value representative of a confidence that the head portion of the person image in the video images is correctly located. The first confidence value and the second confidence value are then outputted. Subsequently, the depiction of the head portion of the person image in the video images is based on the first confidence value and the second confidence value.

Claims

1. A method of processing images to identify a head portion comprising:

obtaining the images;

generating a first confidence value representing a confidence that a first process has identified the head portion in the images;

generating a second confidence value representing a confidence that a second process has identified the head portion in the images; and

identifying the head portion in the images based at least in part on the first confidence value and the second confidence value.

2. A method as recited in claim 1, wherein the first process and the second process employ different processes.

3. A method as recited in claim 1, wherein the first confidence value is based at least in part on information associated with a previously identified head portion.

4. A method as recited in claim 1, wherein the head portion is identified in the images based at least in part on a first head portion identified by the first process and a second head portion identified by the second process.

5. A method as recited in claim 1, wherein the head portion is based at least in part on a previously identified head portion.

6. A method as recited in claim 1, wherein the first process includes analyzing pixels within a search window, wherein the search window is created based on a previously identified head portion.

7. A method as recited in claim 1, wherein the first process includes:

identifying areas as possible head portions; and

ranking the areas based on their similarity to a predefined oval shape.

8. A method as recited in claim 1, wherein the first process includes identifying areas that are greater than a predefined size as possible head portions.

9. A method as recited in claim 1, wherein the first process includes:

identifying areas as possible head portions; and

selecting the head portion from the areas based at least in part on a plurality of factors, wherein a confidence value associated with an area is proportional to the number of factors the area satisfies.

10. A method as recited in claim 1, wherein the first process includes removing a background image from the obtained image.

11. A method as recited in claim 1, wherein the first process includes:

removing a background image from the obtained image to obtain a foreground image; and

identifying a person image in the foreground image.

12. A method as recited in claim 1, wherein the first process includes generating a mass distribution, wherein local maxima of the mass distribution are used to identify a person image.

13. A method as recited in claim 1, wherein the first process includes generating a mass distribution, wherein local maxima of the mass distribution in the context of a previously identified person image are used to identify a person image.

14. A method as recited in claim 1, wherein the first process includes identifying a point of separation between the head portion and a torso portion.

15. A method as recited in claim 1, wherein the first process includes:

in the event that the first confidence value is below a threshold, identifying an initial location of the head portion; and

identifying a current location of the head portion based at least in part on the initial location of the head portion.

16. A method as recited in claim 1, wherein the first process includes:

in the event that the first confidence value is below a threshold, identifying an initial location of the head portion based at least in part on motion detection; and

identifying a current location of the head portion based at least in part on the initial location of the head portion.

17. A method as recited in claim 1, wherein the first process includes:

in the event that the first confidence value is below a threshold, identifying an initial location of the head portion; and

identifying a current location of the head portion based at least in part on the initial location of the head portion and motion detection.

18. A method as recited in claim 1, wherein the first process includes:

in the event that the first confidence value is below a threshold, identifying an initial location of the head portion based at least in part on detection of a skin color; and

identifying a current location of the head portion based at least in part on the initial location of the head portion.

19. A method as recited in claim 1, wherein the first process includes:

in the event that the first confidence value is below a threshold, identifying an initial location of the head portion; and

identifying a current location of the head portion based at least in part on the initial location of the head portion and detection of a skin color.

20. A method as recited in claim 1, wherein the first process includes:

in the event that the first confidence value is below a threshold, identifying an initial location of the head portion; and

identifying a current location of the head portion based at least in part on the initial location of the head portion, including;

generating a three-dimensional color histogram associated with an area within the image; and

comparing the three-dimensional color histogram to a color model.

21. A method as recited in claim 1, wherein the first process includes:

in the event that the first confidence value is below a threshold, identifying an initial location of the head portion; and

identifying a current location of the head portion based at least in part on the initial location of the head portion, including;

generating a motion distribution map associated with an area within the image; and

identifying an area of motion using a motion histogram based on the motion distribution map.

22. A system for processing images to identify a head portion comprising:

an interface configured to obtain the images; and

a processor configured to identify the head portion in the images based at least in part on a first confidence value and a second confidence value.

23. A system as recited in claim 22, wherein:

the first confidence value represents a confidence that a first process has identified the head portion in the images;

the second confidence value represents a confidence that a second process has identified the head portion in the images; and

the first process and the second process employ different processes.

24. A system as recited in claim 22, wherein:

the first confidence value represents a confidence that a first process has identified the head portion in the images;

the second confidence value represents a confidence that a second process has identified the head portion in the images; and

the head portion is identified in the images based at least in part on a first head portion identified by the first process and a second head portion identified by the second process.

25. A system as recited in claim 22, wherein:

the first confidence value represents a confidence that a first process has identified the head portion in the images;

the second confidence value represents a confidence that a second process has identified the head portion in the images; and

the first process includes removing a background image from the obtained image.

26. A system as recited in claim 22, wherein:

the first confidence value represents a confidence that a first process has identified the head portion in the images;

the second confidence value represents a confidence that a second process has identified the head portion in the images; and

the first process includes generating a mass distribution, wherein local maxima of the mass distribution are used to identify a person image.

27. A system as recited in claim 22, wherein:

the first confidence value represents a confidence that a first process has identified the head portion in the images;

the second confidence value represents a confidence that a second process has identified the head portion in the images; and

the first process includes;

in the event that the first confidence value is below a threshold, identifying an initial location of the head portion; and

identifying a current location of the head portion based at least in part on the initial location of the head portion.

28. A computer program product for processing images to identify a head portion, the computer program product being embodied in a computer readable medium and comprising computer instructions for:

obtaining the images;

generating a first confidence value representing a confidence that a first process has identified the head portion in the images;

generating a second confidence value representing a confidence that a second process has identified the head portion in the images; and

identifying the head portion in the images based at least in part on the first confidence value and the second confidence value.

29. A computer program product as recited in claim 28, wherein the first process and the second process employ different processes.

30. A computer program product as recited in claim 28, wherein the head portion is identified in the images based at least in part on a first head portion identified by the first process and a second head portion identified by the second process.

31. A computer program product as recited in claim 28, wherein the first process includes removing a background image from the obtained image.

32. A computer program product as recited in claim 28, wherein the first process includes generating a mass distribution, wherein local maxima of the mass distribution are used to identify a person image.

33. A computer program product as recited in claim 28, wherein the first process includes:

in the event that the first confidence value is below a threshold, identifying an initial location of the head portion; and

identifying a current location of the head portion based at least in part on the initial location of the head portion.