METHOD AND APPARATUS FOR AUTOMATIC ILLUMINANT COMPENSATION IN VIDEO SURVEILLANCE

• US 20100150409A1
• Filed: 12/10/2009
• Published: 06/17/2010
• Est. Priority Date: 12/12/2008
• Status: Active Grant

• Current Assignee: Project Giants, LLC
• Original Assignee: Tektronix Incorporated (Fortive Corp.)
• Inventors: FERGUSON, KEVIN M.
• Granted Patent: US 8,781,222 B2
• Time in Patent Office: Days
• US Class Current: 382/118
• CPC Class Codes: H04N 9/73 Colour balance circuits, e.g. white balance circuits or colour temperature control camera processing pipelines for colour balance H04N23/88

Abstract

Embodiments of the invention include systems and methods for automatically detecting persons and/or other objects in a video signal. Automatic color compensation is applied to the video signal based on measured color of a reference area of the video signal. Generally the reference area is an area of the video signal that does not change often over time. The color compensated video signal is then analyzed for the appearance of objects and persons. In some instances the persons are identified using facial or other type of recognition. Counters or other trackers may provide real-time information, such as how many people are currently on a particular floor of a building.

Claims

1. A method of automatic object detection in video comprising:

receiving a video signal at a video input, the video signal including objects lit under varying lighting conditions;

measuring a color of a reference area within the video signal;

providing the measured color as a color reference input to a color processor; and

adjusting a color signal of data generated from the video signal using the measured color.

2. The method of automatic object detection according to claim 1 in which measuring a color of a reference area within the video signal comprises measuring a white signal of the reference area.

3. The method of automatic object detection according to claim 2 in which measuring a white signal comprises calculating a white value from two or more color separated signals.

4. The method of automatic object detection according to claim 1, further comprising detecting human objects using the adjusted color signal and a human-skin color reference.

5. The method of automatic object detection according to claim 4, in which the human-skin color reference is a plurality of different skin color references.

6. The method of automatic object detection according to claim 1, further comprising classifying human objects using the adjust color signal and a human characteristic database.

7. The method of automatic object detection according to claim 1, further comprising keeping a present count of a number of persons within a building under surveillance.

8. A method of face detection, comprising:

receiving a video signal at a video input, the video signal including one or more human faces lit under varying lighting conditions;

measuring a color of a reference area within the video signal;

providing the measured color as a color reference input to a color processor; and

adjusting a color signal of data generated from the human faces of the video signal using the measured color.

9. The method of face detection according to claim 8 in which measuring a color of a reference area within the video signal comprises measuring a white signal of the reference area.

10. The method of face detection according to claim 9 in which measuring a white signal comprises calculating a white value from two or more color separated signals.

11. The method of face detection according to claim 8, further comprising associating one or more faces from the adjusted color signal with one or more stored facial identities.

12. A system for automatic object detection, comprising:

a video input structured to accept a video signal;

one or more video models structured to generate video data from the video signal;

a color reference generator structured to abstract color data from a pre-selected portion of the video data;

a color processor structured to modify the video data over time using color data abstracted from the color reference generator; and

an object detector coupled to the temporal processor and structured to compare the modified video data against pre-stored object criteria.

13. The system of claim 12 in which the color reference generator produces a different output when ambient lighting within the video signal changes.

14. The system of claim 12 in which the pre-selected portion of the video data is a portion of a wall.

15. The system of claim 12, in which the color reference generator is an automatic color reference generator.

16. The system of claim 15 in which the automatic color reference generator is structured to determine a target area of the video signal that has relatively fewer changes over time as compared to other portions of the video signal, and to use the determined target area as the pre-selected portion of the video data.

17. The system of claim 12 in which the color reference generator is formed from white components of a set of cone responses of the video signal.

18. The system of claim 12 in which the object detector comprises a facial recognition database, and in which the pre-stored object criteria are facial features.

19. The system of claim 12 in which the object detector comprises a skin detector, and in which the pre-stored object criteria include human identifying information.

20. The system of claim 12 in which the object detector comprises an object counter structured to count a number of pre-defined objects that appear on the video signal.