Video scene background maintenance using change detection and classification

• US 20040151342A1
• Filed: 01/30/2003
• Published: 08/05/2004
• Est. Priority Date: 01/30/2003
• Status: Active Grant

• Current Assignee: Motorola Solutions, Inc.
• Original Assignee: ObjectVideo Incorporated (Alarm.com Holdings, Inc.)
• Inventors: Haering, Niels, Chosak, Andrew J., Zhang, Zhong, Venetianer, Peter L., Lipton, Alan J.
• Granted Patent: US 6,999,600 B2
• Time in Patent Office: Days
• US Class Current: 382/103
• CPC Class Codes: G06T 2207/10016 Video; Image sequence
• CPC Class Codes: G06T 7/11 Region-based segmentation
• CPC Class Codes: G06T 7/143 involving probabilistic approaches, e.g. Markov random field [MRF] modelling
• CPC Class Codes: G06T 7/194 involving foreground-background segmentation
• CPC Class Codes: G06T 7/254 involving subtraction of images
• CPC Class Codes: G06V 10/28 Quantising the image, e.g. histogram thresholding for discrimination between background and foreground patterns
• CPC Class Codes: G06V 20/00 Scenes; Scene-specific elements control of digital cameras H04N23/60

Abstract

Video is processed by maintaining a background model for the video, detecting a target in the video, detecting if the target is a stationary target, and classifying the stationary target as an insertion in the background model or a removal from the background model.

Claims

1. A method for processing video, comprising the steps of:

maintaining a background model for said video;

detecting a target in said video;

detecting if said target is a stationary target; and

classifying said stationary target as an insertion in said background model or a removal from said background model.

2. A method as in claim 1, wherein detecting if said target is said stationary target comprises the steps of:

determining general motion and size change properties of said target;

determining independent motion properties of said target; and

determining if said target is said stationary target based on said general motion and size change properties and said independent motion properties.

3. A method as in claim 2, wherein said general motion and size change properties comprise statistical properties of a centroid trajectory of said target and statistical properties of an area of said target.

4. A method as in claim 2, wherein said independent motion properties comprise statistical properties of moving pixels of said target.

5. A method as in claim 2, wherein determining if said target is stationary is further based on aging of said target.

6. A method as in claim 1, wherein classifying said stationary target comprises the steps of:

determining an edge strength in said background model along a boundary of said stationary target;

determining an edge strength in a current frame of video sequence along said boundary of said stationary target; and

determining if said stationary target is said insertion or said removal based on said edge strength in said background model and said edge strength in said current frame.

7. A method as in claim 6, wherein determining said edge strength in said background model is based on a centroid pixel of said stationary target and perimeter pixels along said boundary of said stationary target.

8. A method as in claim 6, wherein determining said edge strength in said current frame is based on a centroid pixel of said stationary target and perimeter pixels along said boundary of said stationary target.

9. A method as in claim 1, wherein classifying said stationary target further comprises classifying said stationary target as an insertion in said background model, a removal from said background model, or as being unidentifiable as said insertion or said removal.

10. A method as in claim 1, further comprising the step of:

determining if said target was previously detected as a stationary target.

11. A computer system comprising a computer-readable medium having software to operate a computer in accordance with the method of claim 1.

12. A computer-readable medium having software to operate a computer in accordance with the method of claim 1.

13. A computer system for processing video, comprising:

a background model of said video;

a background model-based pixel classification to produce a change mask and imagery based on said video and said background model;

a background model update to update said background model based on said change mask and said imagery;

a motion-based pixel classification to produce a motion mask based on said video;

a blob generation to produce at least one blob based on said change mask and said motion mask;

a blob tracking to produce at least one target based on said blobs;

a stationary target detection and classification to produce a stationary target description based on each target, said stationary target description to identify each said target as an insertion in said background model or a removal from said background model; and

a background model local update to update said background model based on each said stationary target description.

14. A computer system as in claim 13, wherein said stationary target descriptions further identify each of said targets as an insertion in said background model, a removal from said background model, or as being unidentifiable as said insertion or said removal.

15. A computer system as in claim 13, further comprising:

a stationary target monitor to produce a target reactivation for said blob tracking based each said stationary target description.

16. An apparatus for processing video, comprising the steps of:

means for maintaining a background model for said video;

means for detecting a target in said video;

means for detecting if said target is a stationary target; and

means for classifying said stationary target as an insertion in said background model or a removal from said background model.