Multiple object processing in wide-angle video camera

• US 7,528,881 B2
• Filed: 04/30/2004
• Issued: 05/05/2009
• Est. Priority Date: 05/02/2003
• Status: Expired due to Fees

• Current Assignee: Grandeye Limited (Oncam Global Group AG)
• Original Assignee: Grandeye Limited (Oncam Global Group AG)
• Inventors: Ahiska, Yavuz
• Primary Examiner(s): Tran; Nhan T
• Assistant Examiner(s): Hsu; Amy
• Application Number:
  US10/837,326
• Publication Number:
  US 20050007479A1
• Time in Patent Office: 1,831 Days
• Field of Search: 348/36, 348/39, 348/143, 348/148, 348/169, 348/14.01, 348/14.02, 348/211.12, 382/115, 382/104, 382/118, 382/105
• US Class Current: 348/335
• CPC Class Codes: H04N 23/62 Control of parameters via user interfaces
• CPC Class Codes: H04N 23/698 for achieving an enlarged field of view, e.g. panoramic image capture
• CPC Class Codes: H04N 23/81 for suppressing or minimising disturbance in the image signal generation

Abstract

A multi object processing video camera is disclosed. The video camera comprises an image scan and capture circuit capable of capturing a wide-angle field of view in high resolution and a processing circuit capable of executing a plurality of software tasks on a plurality of regions from the image sensor and capture circuit based upon the wide angle field of view. The video camera includes memory for storing programs for the processing circuit based upon the captured wide angle field of view and a network connection coupled to the processing circuit. A wide-angle motion video camera for use in surveillance and security applications is disclosed in which multiple objects can be processed by the camera and the results of the processing transmitted to a base station.

Claims

1. A multi object processing video camera, comprising:

a wide-angle optical unit connected to a image sensor and a capture circuit configured to capture a wide-angle field of view in high resolution;

a processing circuit inside the camera which;

executes a plurality of software tasks on a plurality of regions from said image sensor and capture circuit based upon said wide-angle field of view one of said software tasks being always the correction of geometric distortions introduced by said wide-angle optical system;

monitors multiple specific objects without losing sight of other objects, the processing of the digital image covering a wide-angle frame or regions of interest within the frame, and simultaneously able to monitor multiple objects;

selectively conditions transmission of a partial set of images from the camera based on the results of image processing event detection and said action of monitoring multiple objects in ones of software tasks; and

an internal program and data memory to hold said software for execution by said processing circuit.

2. The multi object processing video camera of claim 1 which includes a network connection coupled to said processing circuit.

3. The multi object processing video camera of claim 2 wherein said network connection utilizes an Internet protocol.

4. The multi object processing video camera of claim 2 in which said network connection is used to receive control commands from a base station.

5. The multi object processing video camera of claim 2 in which said network connection is used to load software into said program and data memory.

6. The multi object processing video camera of claim 1 in which said software tasks include tracking of object movement, said object'"'"'s spatial coordinates and or trajectory being transferred to one or more base stations or other cameras.

7. The multi object processing video camera of claim 2 in which said software tasks include smoke detection.

8. The multi object processing video camera of claim 1 in which said software tasks include left package detection.

9. The multi object processing video camera of claim 2 in which said software tasks include face recognition by comparison with data loaded into said camera by said network connection.

10. The multi object processing video camera of claim 2 in which said software tasks include vehicle license plate recognition by comparison with data loaded into said camera by said network connection.

11. The multi object processing video camera of claim 1 in which said software tasks include behavior analysis.

12. The multi object processing video camera of claim 1 including magnetic storage in addition to said program and data memory for the purpose of storing captured video data, said data being either wide-angle or a selected region and either as captured or processed.

13. The multi object processing video camera of claim 12 in which said stored video data is made available for further processing within said camera.

15. The video camera of claim 1 wherein said image processing circuit is a digital signal processor.

16. The video camera of claim 15 wherein said digital signal processor is comprised of an arithmetic and logic unit, an instruction sequencer, an instruction cache, a memory module, a data cache, and a register storage.

17. The video camera of claim 15 wherein said digital signal processor is used to execute algorithms upon an image selected from the group consisting of:

sharpening, contrast adjustment, color adjustment, resizing, motion detection, smoke detection, and abandoned package detection.

18. The video camera of claim 1 wherein said image processing circuit is a programmable graphics processor.

19. The video camera of claim 18 wherein said programmable graphics processor is comprised of a rasterizer, a pixel shader, a memory module, a data cache, a pixel filter, and a pixel write unit.

14. A multi object processing video camera, comprising:

a wide-angle optical unit connected to a image sensor and a capture circuit configured to capture a wide-angle field of view in high resolution;

a processing circuit within the camera which;

executes a plurality of software tasks on a plurality of regions from said image sensor and capture circuit based upon said wide-angle field of view, one of said software tasks being always the correction of geometric distortions introduced by said wide-angle optical system;

monitors multiple specific objects without losing visibility of other objects, the processing of the digital image covering a wide-angle frame or regions of interest within the frame, and simultaneously monitors multiple objects; and

selectively conditions transmission of images from said camera to a base station connected to said camera based on the results of an event detection algorithm and said action of monitoring multiple objects in at least one of said software tasks; and

an internal program and data memory to hold said software for execution by said processing circuit.

21. The multi object processing video camera of claim 14, further comprising:

the processing circuit produces coordinates of an output image and passes them to a coupled pixel shader, which calculates the pixels in the captured digital image to use and reads said pixels from a memory to generate the high resolution image.

20. A multi object processing video camera, comprising:

a wide-angle optical unit connected to a image sensor and a capture circuit configured to capture a wide-angle field of view in high resolution;

a processing circuit within the camera which;

executes a plurality of software tasks on a plurality of regions from said image sensor and capture circuit based upon said wide-angle field of view, one of said software tasks being always the correction of geometric distortions introduced by said wide-angle optical system; and

selectively conditions transmission of a partial set of images from said wide-angle optical unit to a base station connected to said wide-angle optical unit based on;

the results of event detection image processing in at least one of said software tasks; and

a validation of said results by another multi object processing wide-angle optical unit connected to said wide-angle optical unit; and

the wide-angle optical unit including software to autonomously identify a specific object, while maintaining sight of other objects, to track and communicate position and trajectory data of the object to at least one other wide-angle optical unit, or receive position and trajectory data from another wide-angle optical unit, and automatically maintain sight of the specific object as it moves across the field of view of multiple wide-angle optical units; and

a program and data memory to hold said software for execution by said processing circuit.