VIDEO COMPRESSION SYSTEM

• US 20080187047A1
• Filed: 10/16/2007
• Published: 08/07/2008
• Est. Priority Date: 10/17/2006
• Status: Active Grant

• Current Assignee: Harman Becker Automotive Systems GmbH (Samsung Electronics Co. Ltd.)
• Original Assignee: Harman Becker Automotive Systems GmbH (Samsung Electronics Co. Ltd.)
• Inventors: Mohr, Ulrich, Stephan, Martin
• Granted Patent: US 8,358,686 B2
• Time in Patent Office: Days
• US Class Current: 375/240.16
• CPC Class Codes: H04N 19/137 Motion inside a coding unit, e.g. average field, frame or block difference
• CPC Class Codes: H04N 19/527 Global motion vector estimation
• CPC Class Codes: H04N 19/61 in combination with predictive coding

Abstract

A video compression system processes images captured from a video camera mounted to a vehicle. Vehicle-mounted sensors generate vehicle motion information corresponding to a current state of motion of the vehicle. An optical flow estimation circuit estimates apparent motion of objects within a visual field. A video encoder circuit in communication with the optical flow estimation circuit compresses the video data from the video camera based on the estimated apparent motion.

Claims

1. A video compression system comprising:

a video camera mounted to or in a vehicle, and configured to generate video data;

at least one sensor configured to generate vehicle motion information corresponding to a current state of motion of the vehicle;

an optical flow estimation circuit configured to generate an estimated apparent motion of objects within a visual field of the video camera; and

a video encoder circuit in communication with the optical flow estimation circuit, and configured to compress the video data in accordance with the estimated apparent motion.

2. The system of claim 1, where the optical flow estimation circuit comprises:

a motion reconstruction circuit configured to reconstruct the current state of motion of the vehicle;

the reconstruction based on the vehicle motion information provided by the sensor; and

where the estimated apparent motion is determined in accordance with the reconstructed state of motion of the vehicle.

3. The system of claim 1, where the estimated apparent motion includes qualitative information indicating a type of flow field.

4. The system of claim 1, where the estimated apparent motion includes a vector field representation.

5. The system of claim 1, where the video encoder circuit comprises:

a model circuit configured to generate motion vectors indicating motion within the video data, the motion vectors determined based on the estimated apparent motion; and

a predictive coding circuit configured generate predicted video data based on the motion vectors, and configured to determine differences between the video data and predicted video data.

6. The system of claim 5, where the model circuit generates the motion vectors by searching a predefined search range of possible motion vectors.

7. The system of claim 6, where the model circuit determines a search range based on the estimated apparent motion.

8. The system of claim 6, where the model circuit determines a starting value for searching the motion vectors based on the estimated apparent motion.

9. The system of claim 5, where the video encoder circuit comprises a segmentation circuit configured to partition video images into blocks, and where the model circuit determines motion vectors for each block based on the estimated apparent motion and a location of the block within the video image.

10. The system of claim 1, where the video encoder circuit compresses the video data according to an MPEG (Motion Picture Experts Group) format.

11. The system of claim 1, where the at least one sensor comprises a tachometer, an accelerometer, an angular sensor, a distance sensor, a gyroscope, a compass, or a GPS receiver.

12. The system of claim 1, where the vehicle motion information includes velocity, direction of motion, linear acceleration, or radial acceleration.

13. The system of claim 1, further comprising a digital communication bus to facilitate communication between the sensors and the video encoder circuit.

14. The system of claim 13, comprising:

a video decoder configured to decode the compressed video information; and

a display device in communication with the decoder, where the display device is configured to display the decoded compressed video data.

15. The system of claim 13, comprising a recording device in communication with the video encoder circuit through the communication bus, where the recording device is configured to store the compressed video data.

16. A video compression system for vehicles, comprising:

a video camera mounted to or in the vehicle, the camera configured to generate video data;

at least one sensor configured to generate vehicle motion information corresponding to a current state of motion of the vehicle;

an optical flow estimation circuit configured to generate an estimated apparent motion of objects within a visual field of the video camera;

a video encoder circuit in communication with the optical flow estimation circuit, and configured to compress the video data in accordance with the estimated apparent motion; and

the video encoder circuit further comprising

a model circuit configured to generate motion vectors indicating motion within the video data, the motion vectors determined based on the estimated apparent motion; and

a predictive coding circuit configured generate predicted video data based on the motion vectors, and configured to determine differences between the video data and predicted video data.

17. A method for compressing video data generated by a video camera mounded in or on a vehicle, the method comprising:

providing vehicle motion information corresponding to a current state of motion of the vehicle;

estimating an apparent motion of objects within a visual field of the video camera based on the vehicle motion information; and

compressing the video data based on the estimated apparent motion.

18. The method of claim 17, comprising:

providing sensor data corresponding to the vehicle motion information;

reconstructing the current state of motion of the vehicle based on the sensor data; and

where estimating the apparent motion of objects is based on the reconstructed current state of motion of the vehicle.

19. The method of claim 17, where estimating the apparent motion of objects generates qualitative information indicating a type of flow field.

20. The method of claim 17, where estimating the apparent motion of objects generates vector field representation information.

21. The method of claim 17 comprising:

determining motion vectors indicative of motion within a subsequent video image, the motion vectors determined based on the estimated apparent motion of objects within the visual field of the camera; and

predictively coding differences between the video data and video images predicted from the motion vectors.

22. The method of claim 21, comprising determining the motion vectors by searching a predefined search range of possible motion vectors.

23. The method of claim 22, comprising determining the motion vectors by establishing a search range in accordance with the estimated apparent motion.

24. The method of claim 22, comprising:

segmenting a video image into a plurality of data blocks; and

determining motion vectors for each block in accordance with the estimated apparent motion and a location of the block within the video image.

25. A method for compressing video data generated by a video camera mounted in or on a vehicle, the method comprising:

segmenting a video image into a plurality of data blocks;

providing vehicle motion information corresponding to a current state of motion of the vehicle;

reconstructing the current state of motion of the vehicle based on the vehicle motion information;

estimating an apparent motion of objects within a visual field of the video camera based on the vehicle motion information and the reconstructed current state of motion of the vehicle;

determining motion vectors for each block in accordance with the estimated apparent motion and a location of the block within the video image; and

compressing the video data based on the motion vectors.