Method and Apparatus for Detecting People by a Surveillance System

• US 20140139660A1
• Filed: 11/21/2012
• Published: 05/22/2014
• Est. Priority Date: 11/21/2012
• Status: Active Grant

• Current Assignee: Pelco Incorporated (Motorola Solutions, Inc.)
• Original Assignee: Pelco Incorporated (Motorola Solutions, Inc.)
• Inventors: Zhu, Hongwei, Wang, Lei, Aghdasi, Farzin, Millar, Greg M.
• Granted Patent: US 9,367,733 B2
• Time in Patent Office: Days
• US Class Current: 348/143
• CPC Class Codes: G06F 18/24 Classification techniques
• CPC Class Codes: G06V 10/46 Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features colour feature extraction G06V10/56
• CPC Class Codes: G06V 20/52 Surveillance or monitoring of activities, e.g. for recognising suspicious objects recognising microscopic objects G06V20/69
• CPC Class Codes: G06V 40/10 Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
• CPC Class Codes: G06V 40/103 Static body considered as a whole, e.g. static pedestrian or occupant recognition

Abstract

Surveillance systems may be found in both private and public spaces. In private spaces, they can be designed to help provide and monitor secure premises. Similarly, public spaces may also use surveillance systems to determine an allocation of public resources. A camera surveillance system according to an embodiment of the invention uses advanced image processing techniques to determine whether an object moving across a scene is a person. The camera surveillance system achieves an accurate and efficient classification by selectively processing a set of features associated with the object, such as features that define an omega shape. By selectively processing the set of features associated with the object, the methods and systems described herein reduce the computational complexity of standard image processing/object detection techniques.

Claims

1. A method for detecting people by a surveillance system, the method comprising:

identifying a foreground contour shape within a scene and a track of the foreground contour shape from at least two frames in a video file;

determining features of the foreground contour shape within the scene and the track of the foreground contour shape; and

classifying at least a portion of the foreground contour shape using a shape feature and the track of the foreground contour shape to determine whether the foreground contour shape matches a person reference model.

2. The method of claim 1, further comprising:

applying an intra-frame local detector to the foreground contour shape to determine whether the foreground contour shape matches the person reference model.

3. The method of claim 1, further comprising:

identifying a plurality of foreground contour shapes within the scene and a plurality of tracks associated with each one of the plurality of foreground contour shapes in the video file; and

assigning a plurality of intra-frame local detectors to visited scene locations;

applying each one of the assigned plurality of intra-frame detectors to each one of the visited scene locations to determine whether each one of the foreground contour shape matches the person reference model.

4. The method of claim 1, further comprising

extracting features from the foreground object within the scene, the features including at least one of the following an object shape, object size, object height, object width, aspect ratio, directional aspect ratio of salient moving direction, head-shoulder feature, aspect ratio of object shape and histogram for chain codes of object shape.

5. The method of claim 1, further comprising:

identifying chain codes defining a perimeter of the foreground contour shape;

determining at least the object shape and object size from the chain codes; and

classifying the object shape and object size from the chain codes to determine whether the foreground contour shape matches a person reference model.

6. The method of claim 1, further comprising:

determining if the calculated object size is large wherein a large object comprises at least 30% of frame size of the video file.

7. The method of claim 1, further comprising:

extracting salient moving directions from the track of the foreground contour shape in the video file; and

extracting a directional aspect ratio from the track of the foreground contour shape in the video file; and

determining whether the foreground contour shape matches the person reference model based on the salient moving direction and the directional aspect ratio.

8. The method of claim 7, further comprising:

determining expected movement from the track of the foreground contour shape, extracted salient moving directions and directional aspect ratio; and

propagating features of the track of the foreground contour shape.

9. The method of claim 1 wherein determining whether the foreground contour shape matches the person reference model includes detecting a head and shoulder shape.

10. The method of claim 9 wherein detecting a head and shoulder shape includes applying an omega classifier.

11. A computer readable medium, the computer readable medium having program instructions stored thereon, the program instructions being executable by a processor, and, when executed by a processor, cause the processor to:

identify a foreground contour shape within a scene and a track of the foreground contour shape from at least two frames in a video file;

determine features of the foreground contour shape within the scene and the track of the foreground contour shape; and

classify at least a portion of the foreground contour shape using a shape feature and the track of the foreground contour shape to determine whether the foreground contour shape matches a person reference model.

12. The computer readable medium of claim 11, further including program instructions that when executed by the processor, cause the processor to:

applying an intra-frame local detector to the foreground contour shape to determine whether the foreground contour shape matches the person reference model.

13. The computer readable medium of claim 11, further including program instructions that when executed by the processor, cause the processor to:

identify a plurality of foreground contour shapes within the scene and a plurality of tracks associated with each one of the plurality of foreground contour shapes in the video file; and

assign a plurality of intra-frame local detectors to visited scene locations;

apply each one of the assigned plurality of intra-frame detectors to each one of the visited scene locations to determine whether each one of the foreground contour shape matches the person reference model.

14. The computer readable medium of claim 11, further including program instructions that when executed by the processor, cause the processor to:

extract features from the foreground object within the scene, the features including at least one of the following an object shape, object size, object height, object width, aspect ratio, directional aspect ratio of salient moving direction, head-shoulder feature, aspect ratio of object shape and histogram for chain codes of object shape.

15. The computer readable medium of claim 11, further including program instructions that when executed by the processor, cause the processor to:

identify chain codes defining a perimeter of the foreground contour shape;

determine at least the object shape and object size from the chain codes; and

classify the object shape and object size from the chain codes to determine whether the foreground contour shape matches a person reference model.

16. A surveillance system, comprising:

a camera configured to capture image data from a scene;

a scene analyzer in communication with the camera and configured to receive image data of the scene, the scene analyzer further configured to;

identify a foreground contour shape within a scene and a track of the foreground contour shape from at least two frames in a video file;

determine features of the foreground contour shape within the scene and the track of the foreground contour shape;

classify at least a portion of the foreground contour shape using a shape feature and the track of the foreground contour shape to determine whether the foreground contour shape matches a person reference model; and

a display device configured to present analysis results from the scene analyzer through a user interface.

17. The surveillance system of claim 16, wherein the camera is a fixed surveillance camera.

18. The surveillance system of claim 16, wherein the camera is configured to support at least two simultaneous video streams.

19. The surveillance system of claim 18, wherein the at least two simultaneous bit streams are customizable.

20. The surveillance system of claim 18, wherein the scene analyzer is configured to:

customize a high bit rate stream according to intra-frame local detector instructions; and

apply the intra-frame local detector to the foreground contour shape to determine whether the foreground contour shape matches the person reference model.