Model-based incident detection system with motion classification

• US 6,985,172 B1
• Filed: 01/25/2002
• Issued: 01/10/2006
• Est. Priority Date: 12/01/1995
• Status: Expired due to Term

• Current Assignee: Southwest Research Institute
• Original Assignee: Southwest Research Institute
• Inventors: Franke, Ernest A., Rigney, Michael P., Magee, Michael J.
• Primary Examiner(s): Dastouri, Mehrdad
• Assistant Examiner(s): REKSTAD, ERICK J
• Application Number:
  US10/056,665
• Time in Patent Office: 1,446 Days
• Field of Search: 348/143, 348/152, 348/169, 348/149, 382/103, 382/262, 382/291, 382/118, 375/240.08
• US Class Current: 348/149
• CPC Class Codes: G06T 2207/30241 Trajectory
• CPC Class Codes: G06T 7/246 using feature-based methods, e.g. the tracking of corners or segments
• CPC Class Codes: G06T 7/254 involving subtraction of images
• CPC Class Codes: G06V 20/54 of traffic, e.g. cars on the road, trains or boats
• CPC Class Codes: G06V 40/20 Movements or behaviour, e.g. gesture recognition recognition of facial expressions G06V40/16
• CPC Class Codes: G08B 13/19602 Image analysis to detect motion of the intruder, e.g. by frame subtraction
• CPC Class Codes: G08B 13/19604 involving reference image or background adaptation with time to compensate for changing conditions, e.g. reference image update on detection of light level change
• CPC Class Codes: G08B 13/19606 Discriminating between target movement or movement in an area of interest and other non-signicative movements, e.g. target movements induced by camera shake or movements of pets, falling leaves, rotating fan
• CPC Class Codes: G08B 13/19613 Recognition of a predetermined image pattern or behaviour pattern indicating theft or intrusion
• CPC Class Codes: G08B 13/19691 Signalling events for better perception by user, e.g. indicating alarms by making display brighter, adding text, creating a sound

Abstract

Surveillance apparatus and methods effective for detecting incidents are based upon improved image processing techniques applied to infrared and visible light spectrum images in a time sequence. A reference image is generated by removing motion objects from an image of a region of interest. The reference image is compared to an image to be analyzed to detect motion. Detected motion is analyzed and classified as motion objects. The motion objects are then compared to a motion model, which classifies the objects as being anticipated or unanticipated.

Claims

1. A method of using video images to monitor incidents in a region of interest, comprising the steps of:

computing a reference image from a set of images by removing moving objects from the set of images;

storing a motion model representing anticipated motion in the region of interest;

acquiring a image to be analyzed;

computing a temporal difference image by comparing the image to be analyzed with the reference image;

repeating the step of computing a temporal difference image to obtain a set of temporal difference images;

calculating, from the set of temporal difference images, at least one temporal difference statistic for each pixel;

detecting motion in a temporal difference image by separating motion pixels from background in the temporal difference image, using the temporal difference statistic;

grouping motion pixels into motion objects;

extracting features from the objects;

comparing the features of motion objects to the motion model;

wherein the motion model classifies objects as anticipated or not anticipated; and

wherein the motion model further classifies anticipated objects as being of interest or not of interest.

2. The method of claim 1, wherein the step of computing a reference image is performed by calculating a median pixel value for each pixel in an image sequence, the image sequence comprising at least a first image and a second image, and by removing moving objects from the image sequence by replacing each pixel value in at least one image of the image sequence with the median pixel value.

3. The method of claim 1, wherein the step of computing a reference image is performed by determining a minimum pixel value for each pixel in an image sequence, the image sequence comprising at least a first image and a second image, and by removing moving objects from the image sequence by replacing each pixel value in at least one image of the image sequence with the minimum pixel value.

4. The method of claim 1, wherein the step of computing a reference image is performed by determining a maximum pixel value for each pixel in an image sequence, the image sequence comprising at least a first image and a second image, and by removing moving objects from the image sequence by replacing each pixel value in at least one image of the image sequence with the maximum pixel value.

5. The method of claim 1, wherein the step of computing a temporal difference image is performed by subtracting pixel values in the image to be analyzed from pixel values of the reference image.

6. The method of claim 1, wherein the step of computing a temporal difference image is performed by subtracting pixel values in the reference image from pixel values of the image to be analyzed.

7. The method of claim 1, wherein the step of computing a temporal difference image is performed by calculating the absolute value of the difference between pixel values in the image to be analyzed and pixel values of the reference image.

8. The method of claim 1, wherein the temporal difference statistic is an nth percentile statistic.

9. The method of claim 1, wherein the temporal difference statistic is a mean statistic.

10. The method of claim 1, wherein the temporal difference statistic is a standard deviation statistic.

11. The method of claim 1, wherein the step of detecting motion is performed by using the temporal difference statistic of each pixel to determine a threshold and by comparing the temporal difference image with the threshold pixel values.

12. The method of claim 11, wherein the threshold is based on an nth percentile statistic.

13. The method of claim 11, wherein the threshold is based on a standard deviation statistic.

14. The method of claim 1, wherein the classifying step is performed by extracting spatial features.

15. The method of claim 1, wherein the classifying step is performed by extracting temporal features.

16. The method of claim 1, further comprising the step of generating a response signal in response to the classifying step.