Scanning camera-based video surveillance system

Patent metadata

Field	Value
Publication number	US-9363487-B2
Application number	US-22097005-A
Country	US
Kind code	B2
Filing date	Sep 8, 2005
Priority date	Sep 8, 2005
Publication date	Jun 7, 2016
Grant date	Jun 7, 2016

Abstract

A video surveillance system may include at least one sensing unit capable of being operated in a scanning mode and a video processing unit coupled to the sensing unit, the video processing unit to receive and process image data from the sensing unit and to detect scene events and target activity.

First claim

We claim:

1. A video surveillance system comprising:
at least one sensing unit operable in a scanning mode that obtains a first view of a scene, an acquisition mode that obtains a second view of the scene, and an interrogation mode; and
a video processing unit, coupled to the at least one sensing unit, that:
obtains video primitives from tracked and classified targets in the first view of the scene by processing image data received from the at least one sensing unit operating in the scanning mode;
detects a scene event or target activity in the first view of the scene by comparing the video primitives to at least one scene event rule or at least one target activity rule;
upon detection, depending on the scene event or target activity detected, automatically switches said at least one sensing unit from the scanning mode to the acquisition mode, wherein the second view of the scene obtained by the acquisition mode provides a closer view, at a variable level of zoom, of the detected scene event or target activity than said first view obtained by the scanning mode;
upon an occurrence of a return event of the acquisition mode, automatically switches the at least one sensing unit from the acquisition mode to the scanning mode;
upon an occurrence of an interrogation event of the acquisition mode, automatically switches said at least one sensing unit from the acquisition mode to the interrogation mode;
analyzes, using the interrogation mode, the detected scene event or target activity; and
upon an occurrence of a return event of the interrogation mode, automatically switches the at least one sensing unit from the interrogation mode.

2. The video surveillance system according to claim 1 , wherein said video processing unit includes:
a vision module that receives video input from said sensing unit;
an inference module that receives output from said vision module; and
a response module that receives output from said inference module.

3. The video surveillance system according to claim 1 , wherein:
the at least one sensing unit comprises a plurality of sensing units;
a first sensing unit of the plurality of sensing units is used in the scanning mode;
a second sensing unit of the plurality of sensing units tracks the scene event or the target activity during the acquisition mode and the interrogation mode; and
the first sensing unit is different than in the second sensing unit.

4. A method, to be performed by video processing equipment, of processing video obtained by a sensing unit, comprising:
scanning, by the sensing unit in a scanning mode, a scene for a target of interest;
obtaining, by processing image data recorded by the sensing unit in the scanning mode, video primitives from tracked and classified targets in the scene; and
detecting a scene event or target activity in a first view of the scene by comparing the video primitives to at least one scene event rule or at least one target activity rule;
upon detecting the scene event or target activity, depending on the scene event or target activity detected in the image data, automatically switching the sensing unit to an acquisition mode that obtains a second view of the scene that provides a closer view, at a variable level of zoom, of the detected scene event or target activity than said first view;
upon an occurrence of a return event of the acquisition mode, automatically switching the sensing unit from the acquisition mode to the scanning mode;
upon an occurrence of an interrogation event of the acquisition mode, automatically switching the sensing unit from the acquisition mode to an interrogation mode that further analyzes the detected scene event or target activity; and
upon an occurrence of a return event of the interrogation mode, automatically switching the sensing unit from the interrogation mode.

5. The method according to claim 4 , wherein said processing image data recorded by the sensing unit in a scanning mode comprises:
identifying the presence, if any, of a target in the first view; and
generating at least one primitive to describe an action performed by the target.

6. The method according to claim 5 , wherein said identifying the presence of a target comprises:
estimating and compensating for motion of said sensing unit.

7. The method according to claim 5 , wherein said identifying the presence of a target comprises:
segmenting a video image corresponding to said first view into foreground and background components;
blobizing the foreground components to create one or more blobs; and
identifying one or more possible targets among said one or more blobs.

8. The method according to claim 7 , wherein said identifying one or more possible targets comprises:
tracking said one or more possible targets; and
classifying said one or more targets.

9. The method according to claim 5 , wherein said processing image data recorded by the sensing unit in a scanning mode includes utilizing at least one scene model to perform said identifying or said generating.

10. The method according to claim 4 , wherein in said acquisition mode, the sensing unit performs operations comprising:
estimating scene motion; and
compensating for motion of the sensing unit being operated in the acquisition mode.

11. The method according to claim 4 , wherein in said acquisition mode, the sensing unit performs operations comprising:
segmenting a video image corresponding to said first view into foreground and background components;
performing template matching on at least one foreground component; and
generating a primitive based on said template matching.

12. The method according to claim 11 , wherein said template matching utilizes an estimate of scene motion.

13. The method according to claim 11 , wherein said template matching utilizes at least one detected image feature.

14. The method according to claim 11 , wherein said template matching utilizes a target model.

15. The method according to claim 11 , further comprising utilizing a scene model to perform at least one activity selected from the group consisting of:
said segmenting a video image; and
said performing template matching; and said generating a primitive.

16. The method according to claim 15 , further comprising:
detecting at least one image feature based on a scene model.

17. The method according to claim 16 , wherein said detecting employs a target model.

18. The method according to claim 4 , wherein in the interrogation mode, the sensing unit performs operations comprising:
segmenting a video image corresponding to said first view into foreground and background components;
identifying and tracking the target; and
generating a primitive based on said identifying and tracking.

19. The method according to claim 18 , wherein identifying and tracking the target comprises:
performing blobization on said foreground components;
tracking said target based on said blobization; and
classifying said target.

20. The method according to claim 18 , wherein identifying and tracking the target comprises:
performing template matching on at least one foreground component; and
wherein said generating a primitive is based on said template matching.

21. The method according to claim 18 , wherein said interrogation mode comprises utilizing a scene model to perform at least one activity selected from the group consisting of: said segmenting a video image; said identifying and tracking the target; and said generating a primitive.

Assignees

• Chosak Andrew J
• Brewer Paul C
• Egnal Geoffrey
• Gupta Himaanshu
• Haering Niels
• Lipton Alan J
• Yu Li
• Avigilon Fortress Corp

Inventors

• Chosak Andrew J
• Brewer Paul C
• Egnal Geoffrey
• Gupta Himaanshu
• Haering Niels
• Lipton Alan J
• Yu Li

Classifications

• H04N23/69

  Control of means for changing angle of the field of view, e.g. optical zoom objectives or electronic zooming · CPC title

• G08B13/19656

  Network used to communicate with a camera, e.g. WAN, LAN, Internet · CPC title

• G08B13/19641

  Multiple cameras having overlapping views on a single scene · CPC title

• G06K9/00771

  Physics · mapped topic

• H04N7/181 Primary

  for receiving images from a plurality of remote sources · CPC title