Network Surveillance System

• US 20100067801A1
• Filed: 11/20/2007
• Published: 03/18/2010
• Est. Priority Date: 11/20/2006
• Status: Active Grant

• Current Assignee: SenSen Networks Ltd.
• Original Assignee: Adelaide Research & Innovation Pty Ltd. (University of Adelaide)
• Inventors: Dick, Anthony Robert, Brooks, Michael John, van den Hengel, Anton John
• Granted Patent: US 8,396,250 B2
• Time in Patent Office: Days
• US Class Current: 382/190
• CPC Class Codes: G06V 10/62 relating to a temporal dimension, e.g. time-based feature extraction; Pattern tracking
• 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: G08B 13/19645 Multiple cameras, each having view on one of a plurality of scenes, e.g. multiple cameras for multi-room surveillance or for tracking an object by view hand-over
• CPC Class Codes: G08B 13/19656 Network used to communicate with a camera, e.g. WAN, LAN, Internet
• CPC Class Codes: H04N 7/18 Closed-circuit television [CCTV] systems, i.e. systems in which the video signal is not broadcast

Abstract

A method for estimating the activity topology of a set of sensed data windows is described. Each of the sensed data windows related to a corresponding sensed region. The method includes the steps of determining a subset of sensed data windows that are not connected; and excluding the subset of sensed data windows from the set of sensed data windows. In one embodiment, the sensed data windows corresponding to image windows such as would be provided by a visual surveillance system.

Claims

1. A method for estimating the activity topology of a set of sensed data windows, each of the sensed data windows related to a corresponding sensed region, the method including the steps:

determining a subset of sensed data windows that are not connected; and

excluding the subset of sensed data windows from the set of sensed data windows.

2. The method of claim 1, wherein the step of determining the subset of sensed data windows includes comparing sensed data windows pair wise from the set of sensed data windows.

3. The method of claim 2, wherein the step of comparing sensed data windows pair wise includes determining whether a pair of sensed data windows does not have overlapping sensed regions.

4. The method of claim 3, wherein the step of determining whether the pair of sensed data windows does not have overlapping sensed regions includes determining a first occupancy measure for a first sensed data window and a second occupancy measure for a second sensed data window, the first and second sensed data windows comprising the pair of sensed data windows, and comparing the first and second occupancy measures.

5. The method of claim 4, wherein the step of comparing the first and second occupancy measures includes forming an occupancy vector corresponding to a first sensed data window sequence associated with the first sensed data window and a second occupancy vector corresponding to a second sensed data window sequence associated with the second sensed data window and comparing corresponding elements of the first and second occupancy vectors.

6. The method of claim 5, wherein the step of comparing corresponding elements of the first and second occupancy vectors includes performing a vector exclusive-or operation on the first and second occupancy vectors to determine that the associated first and second sensed data windows do not overlap.

7. The method of claim 5, wherein the second occupancy vector that corresponds to the second sensed data window sequence is based on the second sensed data window and its nearest neighbouring sensed data windows thereby forming a padded occupancy vector.

8. The method of claim 7, wherein the step of comparing corresponding elements of the first and second occupancy vectors includes performing a vector exclusive-or operation on the first occupancy vector and the padded occupancy vector to determine that the associated first and second sensed data windows do not overlap.

9. The method of claim 7, wherein the step of comparing corresponding elements of the first and second occupancy vectors includes performing a vector uni-directional exclusive-or operation on the first occupancy vector and the padded occupancy vector to determine that the associated first and second sensed data windows do not overlap.

10. The method of claim 5, wherein the step of comparing corresponding elements of the first and second occupancy vectors includes comparing over neighbouring elements of one or both of the first and second occupancy vectors.

11. The method of claim 3, wherein the step of determining whether a pair of sensed data windows does not have overlapping sensed regions includes taking into account the likelihood of a false indication that a pair of sensed data windows do overlap.

12. The method of claim 11, wherein the step of taking into account the likelihood of a false indication that a pair of sensed data windows do overlap is based on previous data associated with the pair of sensed data windows.

13. The method of claim 11, wherein the step of taking into account the likelihood of a false indication that a pair of sensed data windows do overlap includes taking into account an error rate of a sensor or sensors associated with the pair of sensed data windows.

14. The method of claim 5, wherein the sensed data window is an image window and the first and second sensed data window sequences correspond to first and second image window sequences.

15. The method of claim 14, wherein the first and second image window sequences correspond to respective time series of images.

16. The method of claim 15, wherein the time series of images is provided by cameras in a network surveillance system.

17. The method of claim 1, further including the step of determining a further subset of sensor data windows that are connected.

21. The method of claim 12, wherein the step of taking into account the likelihood of a false indication that a pair of sensed data windows do overlap includes taking into account an error rate of a sensor or sensors associated with the pair of sensed data windows.

18. A network surveillance system including:

a network of sensors, each sensor providing one or more sensed data windows each corresponding to a sensed region and forming in total a set of sensed data windows; and

data processing means to determine a subset of sensed data windows that are not connected and excluding the subset of sensed data windows from the set of sensed data windows.

19. A program storage device readable by machine, tangibly embodying a program of instructions to perform method steps for estimating the activity topology of a set of sensed data windows, each of the sensed data windows related to a corresponding sensed region, the method steps including:

determining a subset of sensed data windows that are not connected; and

excluding the subset of sensed data windows from the set of sensed data windows.

20. (canceled)