Method of compression for wide angle digital video

• US 7,894,531 B1
• Filed: 02/15/2006
• Issued: 02/22/2011
• Est. Priority Date: 02/15/2005
• Status: Active Grant

• Current Assignee: Grandeye Limited (Oncam Global Group AG)
• Original Assignee: Grandeye Limited (Oncam Global Group AG)
• Inventors: Cuce, Halil I., Castellari, Andrea Elvis, Mulayim, Adem, Davey, Mark Kenneth, Cetin, Ahmet Enis
• Primary Examiner(s): VO, DON NGUYEN
• Application Number:
  US11/355,262
• Time in Patent Office: 1,833 Days
• Field of Search: 375/240.1, 375/240.03, 375/240.13, 375/240.16, 375/240.18, 375/240.19, 375/240.2, 382/232, 382/238, 382/248, 382/250
• US Class Current: 375/240.19
• CPC Class Codes: H04N 19/12 Selection from among a plurality of transforms or standards, e.g. selection between discrete cosine transform [DCT] and sub-band transform or selection between H.263 and H.264
• CPC Class Codes: H04N 19/17 the unit being an image region, e.g. an object
• CPC Class Codes: H04N 19/176 the region being a block, e.g. a macroblock
• CPC Class Codes: H04N 19/61 in combination with predictive coding
• CPC Class Codes: H04N 19/63 using sub-band based transform, e.g. wavelets

Abstract

A system and method for varying the level of detail encoded in a video. In a preferred embodiment, some regions of a wide-angle video scene are encoded in an almost lossless manner, while other regions are encoded with less detail. Regions of interest can be determined in many ways, including a priori, automatically in real time, or by the selection of human operators.

Claims

1. A method of intelligently compressing a stream of wide angle video images by applying varying compression methodology to selected regions of interest (RoI), comprising the steps of:

using a detection algorithm for determining a plurality of regions of interest in an image frame of the video,

encoding the regions of interest of video using only intra-frame information by computing a linear transform of the image pixels to generate transform domain coefficients, quantizing the transform domain coefficients and representing the quantized coefficients in binary form,

compressing the exterior regions of interest using plurality of image frames of the video, and

combining said compressed exterior regions of interest with said encoded regions of interest to create a bitstream representing the plurality of image frames of the video.

2. The method of claim 1 in which said detection algorithm is a human detection algorithm in wide angle video.

3. The method of claim 1 in which said detection algorithm is implemented on the view corrected video.

4. The method of claim 1 in which the linear transformation used in compresion includes wavelet and the discrete cosine transforms.

6. The method of claim 1 in which said regions of interest are compressed by using intraframe information by canceling the motion estimation and the camera motion compensation process in differentially compressed video data.

7. The method of claim 1 in which differential video data compression methods include a compression standard selected from the group consisting of:

MPEG-2 video compression and MPEG-4 video compression.

8. The method of claim 1 in which said regions of interest in the video are also determined manually by a user.

9. The method of claim 1 in which said regions of interest are determined on the view corrected video.

10. The method of claim 1 in which no side information describing regions of interest is transmitted to the encoder.

5. The method of clim 1 in which the quantized transform domain coefficients are encoded in binary form using coding selected from the group consisting of:

Huffman coding and arthmetic coding.

11. A computer readable medium containing programming instructions for intelligently compressing stream of video image frames by applying varying compressions methodology to selected regions of interest (RoI), comprising the steps of

using a detection algorithm for determining a plurality of regions of interest in an image frame of the video,

encoding the regions of interest of video using only intra-frame information by computing a linear transform of the image pixels to generate transform domain coefficients, quantizing the transform domain coefficients using lower quantization steps for transform domain coefficients and representing the said quantized coefficients in binary form,

compressing the exterior of regions of interests using a plurality of image framed of the video, and

combining the compressed exterior of RoIs with encoded RoIs to create a bitstream representing the plurality of image frames of the video.

12. The computer-readable medium of claim 11 in which the detection algorithm is a human detection algorithm in wide angle video.

13. The computer-readable medium of claim 11 in which said detection algorithm is implemented on the view corrected video.

14. The computer-readable medium of claim 11, wherein the linear transform is selected from the group consisting of the Discrete Cosine Transform and a wavelet transform.

15. The computer-readable medium of claim 11, wherein the quantized transform domain coefficients are encoded in binary form using coding selected from the group of Huffman coding and arithmetic coding.

16. The computer-readable medium of claim 11, in which said regions of interest are compressed by using intraframe information by cancelling the motion estimation and the camera compensation process in differentially compressed video data.

17. The computer-readable medium of claim 11 in which video data compression methods used are selected from the group consisting of MPEG-2 and MPEG-4 video compression standards.

18. The computer-readable medium of claim 11 in which said regions of interests in the video are also determined manually by a user.

19. The computer-readable medium of claim 11 in which said regions of interest are determined on the view corrected video.

20. The computer-readable medium of claim 11 in which no side information describing said Regions of Interest is transmitted to the encoder.