SYSTEM FOR AND METHOD OF SELECTIVE VIDEO FRAME COMPRESSION AND DECOMPRESSION FOR EFFICIENT EVENT-DRIVEN SEARCHING IN LARGE DATABASES

• US 20130101171A1
• Filed: 10/20/2011
• Published: 04/25/2013
• Est. Priority Date: 10/20/2011
• Status: Active Grant

• Current Assignee: Conduent Business Services, LLC (Conduent, Inc.)
• Original Assignee: Xerox Corporation (Xerox Holdings Corp.)
• Inventors: Bulan, Orhan, Bernal, Edgar A., Loce, Robert, Wu, Wencheng
• Granted Patent: US 8,855,436 B2
• Time in Patent Office: Days
• US Class Current: 382/104
• CPC Class Codes: G06V 20/54 of traffic, e.g. cars on the road, trains or boats
• CPC Class Codes: G08G 1/0133 for classifying traffic situation
• CPC Class Codes: H04N 19/107 between spatial and temporal predictive coding, e.g. picture refresh
• CPC Class Codes: H04N 19/139 Analysis of motion vectors, e.g. their magnitude, direction, variance or reliability
• CPC Class Codes: H04N 19/172 the region being a picture, frame or field
• CPC Class Codes: H04N 19/513 Processing of motion vectors
• CPC Class Codes: H04N 19/61 in combination with predictive coding

Abstract

A method of, and system for, selective video frame compression and decompression is presented. The disclosed technique is particularly suited for efficient event-driven searching in large databases. The technique involves designating certain frames in a video sequence as reference frames, to be compressed using a first compression technique, where the remaining frames are compressed using a second compression technique. The first compression technique may be implemented without requiring data from any neighboring frames.

Claims

1. A method of selective video frame compression for object event driven searching, the method comprising:

acquiring electronically stored decompressed video data depicting at least one object in motion, the video data representing a plurality of frames;

detecting frames depicting a desired object position;

compressing each frame depicting a desired object position using a first computer implemented compression technique that does not require other frames in the video data for compression;

compressing a plurality of frames not depicting a desired object position using a second computer implemented compression technique that requires other frames in the video data for compression;

storing compressed video data for future investigation.

2. The method of claim 1 wherein the object one of a motor vehicle, a human-propelled vehicle, a person, an animal, and a product in an assembly line.

3. The method of claim 1 wherein the detecting frames comprises imposing a virtual geometric figure on each frame and performing an action selected from the group consisting of:

(1) detecting a motion vector of sufficient magnitude on the virtual geometric figure in a sufficient number of frames, (2) performing a background subtraction and detecting sufficient movement of an object on the virtual geometric figure, and (3) implementing an object tracking technique to detect sufficient movement of an object on the virtual geometric figure.

4. The method of claim 3 wherein the virtual geometric figure comprises one of a virtual line, a virtual freehand shape and a virtual polygon.

5. The method of claim 1 further comprising accessing compressed video data and decompressing only those frames that were compressed using the first computer implemented compression technique.

6. A system for selective video frame compression for object event driven searching, the system comprising:

a persistent storage comprising electronically stored decompressed video data depicting at least one object in motion, the video data representing a plurality of frames;

an electronic processor configured to detect frames depicting a desired object position;

an electronic processor configured to compress each frame depicting a desired object position using a first computer implemented compression technique that does not require other frames in the video data for compression;

an electronic processor configured to compress a plurality of frames not depicting a desired object position using a second computer implemented compression technique that requires other frames in the video data for compression; and

a persistent storage comprising electronically stored compressed video.

7. The system of claim 6 wherein the object is one of a motor vehicle, a human-propelled vehicle, a person, and a product in an assembly line.

8. The system of claim 6 further comprising an electronic processor configured to impose a virtual geometric figure on each frame and perform an action selected from the group consisting of:

(1) detect a motion vector of sufficient magnitude on the virtual geometric figure in a sufficient number of frames, (2) perform a background subtraction and detect sufficient movement of an object on the virtual geometric figure, and (3) implement an object tracking technique to detect sufficient movement of an object on the virtual geometric figure.

9. The system of claim 8 wherein the virtual geometric figure comprises one of a virtual line, a virtual freehand shape and a virtual polygon.

10. The system of claim 6 further comprising an electronic processor configured to access compressed video data and decompress only those frames that were compressed using the first computer implemented compression technique.

11. A method of searching for a particular object depicted in stored video data, the method comprising:

acquiring electronically stored compressed video data depicting at least one object in motion, the video data, representing a plurality of frames, the compressed video data comprising frames compressed using a first computer implemented compression technique that does not require other frames in the video data for compression and frames compressed using a second computer implemented compression technique that requires other frames in the video data for compression;

decompressing only those frames that were compressed using the first computer implemented compression technique, resulting in a plurality of decompressed frames; and

searching for the particular object in the decompressed frames.

12. The method of claim 11 wherein the object is one of a motor vehicle, a human-propelled vehicle, a person, and a product in an assembly line.

13. The method of claim 11 wherein the frames compressed using the first computer implemented compression technique comprise frames selected from the group consisting of:

(I) frames comprising a motion vector of sufficient magnitude on a virtual geometric figure, (2) frames depicting sufficient movement on a virtual geometric figure after a background subtraction and (3) frames depicting a tracked object exhibiting sufficient movement on a virtual geometric figure.

14. The method of claim 12 wherein the virtual geometric figure comprises one of a virtual line, a virtual freehand shape, and a virtual polygon.

15. The method of claim 11 wherein the searching comprises searching for at least a portion of a license plate number.

16. A system for searching for a particular object depicted in stored video data, the system comprising:

a persistent storage comprising electronically stored compressed video data depicting at least one object in motion, the video data representing a plurality of frames, the compressed video data comprising frames compressed using a first computer implemented compression technique that does not require other frames in the video data for compression and frames compressed using a second computer implemented compression technique that requires other frames in the video data for compression;

an electronic processor configured to decompress only those frames that were compressed using the first computer implemented compression technique, resulting in a plurality of decompressed frames; and

an electronic processor configured to search for the particular object in the decompressed frames.

17. The system of claim 16 wherein the object is one of a motor vehicle, a human-propelled vehicle, a person, and a product in an assembly line.

18. The system of claim 16 wherein the frames compressed using the first computer implemented compression technique comprise frames selected from the group consisting of:

(1) frames comprising a motion vector of sufficient magnitude on a virtual geometric figure, (2) frames depicting sufficient movement on a virtual geometric figure after a background subtraction and (3) frames depicting a tracked object exhibiting sufficient movement on a virtual geometric figure.

19. The system of claim 18 wherein the virtual geometric figure comprises one of a virtual line, a virtual freehand shape, and a virtual polygon.

20. The system of claim 16 wherein the electronic processor configured to search for the particular object in the decompressed frames is further configured to search for at least a portion of a license plate number.