Low power surveillance camera system for intruder detection

Patent metadata

Field	Value
Publication number	US-9544550-B1
Application number	US-201414209136-A
Country	US
Kind code	B1
Filing date	Mar 13, 2014
Priority date	Mar 14, 2013
Publication date	Jan 10, 2017
Grant date	Jan 10, 2017

Abstract

Described is a low power surveillance camera system for intruder detection. The system observes a scene with a known camera motion to generate images with various viewing angles. Next, a background learning mode is employed to generate a low rank matrix for the background in the images. Background null space projections are then learned, which provide a foreground detection kernel. A new scene with known viewing angles is then obtained. Based on the foreground detection kernel and the new input image frame, low power foreground detection is performed to detect foreground potential regions of interest (ROIs), such as intruders. To filter out minimal foreground activity, the system identifies contiguous ROIs to generate the foreground ROI. Focus measures are then employed on the ROIs using foveated compressed sensing to generate foveated measurements. Based on the foveated measurements, the foreground is reconstructed for presentation to a user.

First claim

What is claimed is:

1. A low power surveillance system, the system comprising:
one or more processors and a memory, the memory having executable instructions encoded thereon, such that upon execution of the instructions, the one or more processors performs operations of:
receiving a series of frames from a camera, each frame having a background and a foreground;
generating a background template from the series of frames;
receiving a new image frame of the scene, the new image frame having a background and a foreground;
detecting an intruder image in the foreground of the new image frame;
combining the intruder image with the background template to reconstruct the foreground; and
wherein in detecting an intruder image in the foreground, the foreground is detected using a convex optimization, as follows:

min
e

⁢



Φ
⁢

⁢
e

-
y



2

+

α
⁢


e


1

,

where Φ denotes a measurement matrix, and where e denotes a sparse component in the new image frame d, where y=Φd, and where α denotes a low rank back round.

2. The system as set forth in claim 1 , wherein detecting an intruder image in the foreground of the new image frame further comprises operations of:
detecting potential regions of interest (ROI) in the foreground of the new image frame;
identifying contiguous ROI in the potential ROI; and
reconstructing the ROI using foveated compressive sensing to generate the intruder image.

3. The system as set forth in claim 2 , wherein in receiving a series of frames from a camera, the series of frames include a collection of images with various known viewing angles.

4. The system as set forth in claim 3 , wherein in generating a background template from the series of frames, each frame is stacked as a column vector of data matrix D, which is decomposed to D=A+E, where A is a low-rank matrix that represents the background and E is a sparse matrix that represents the foreground.

5. The system as set forth in claim 1 , wherein in receiving a series of frames from a camera, the series of frames include a collection of images with various known viewing angles.

6. The system as set forth in claim 1 , wherein in generating a background template from the series of frames, each frame is stacked as a column vector of data matrix D, which is decomposed to D=A+E, where A is a low-rank matrix that represents the background and E is a sparse matrix that represents the foreground.

7. A computer program product stored on the non-transitory computer-readable medium for a low power surveillance system, the computer program product comprising:
a non-transitory computer-readable medium having executable instructions, such that upon execution of the instructions by one or more processors, the one or more processors perform operations of:
receiving a series of frames from a camera, each frame having a background and a foreground;
generating a background template from the series of frames;
receiving a new image frame of the scene, the new image frame having a background and a foreground;
detecting an intruder image in the foreground of the new image frame;
combining the intruder image with the background template to reconstruct the foreground; and
wherein in detecting an intruder image in the foreground, the foreground is detected using a convex optimization, as follows:

min
e

⁢



Φ
⁢

⁢
e

-
y



2

+

α
⁢


e


1

,

where Φ denotes a measurement matrix, and where e denotes a sparse component in the new image frame d, where y=Φd, and where α denotes a low rank background.

8. The computer program product as set forth in claim 7 , wherein detecting an intruder image in the foreground of the new image frame further comprises operations of:
detecting potential regions of interest (ROI) in the foreground of the new image frame;
identifying contiguous ROI in the potential ROI; and
reconstructing the ROI using foveated compressive sensing to generate the intruder image.

9. The computer program product as set forth in claim 8 , wherein in receiving a series of frames from a camera, the series of frames include a collection of images with various known viewing angles.

10. The computer program product as set forth in claim 9 , wherein in generating a background template from the series of frames, each frame is stacked as a column vector of data matrix D, which is decomposed to D=A+E, where A is a low-rank matrix that represents the background and E is a sparse matrix that represents the foreground.

11. The computer program product as set forth in claim 7 , wherein in receiving a series of frames from a camera, the series of frames include a collection of images with various known viewing angles.

12. The computer progr

Assignees

• Hrl Lab Llc

Inventors

• Ni Kang-Yu
• Rao Shankar R
• Owechko Yuri

Classifications

• G06V10/7715

  Feature extraction, e.g. by transforming the feature space, e.g. multi-dimensional scaling [MDS]; Mappings, e.g. subspace methods · CPC title

• G06F18/2136

  based on sparsity criteria, e.g. with an overcomplete basis · CPC title

• G06V20/52 Primary

  Surveillance or monitoring of activities, e.g. for recognising suspicious objects (recognising microscopic objects G06V20/69) · CPC title

• H04N7/183 Primary

  for receiving images from a single remote source · CPC title

• G08B13/19604

  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 title