PRIORITIZED COMPRESSION FOR VIDEO

• US 20130121563A1
• Filed: 11/13/2012
• Published: 05/16/2013
• Est. Priority Date: 11/14/2011
• Status: Active Grant

• Current Assignee: NVIDIA Corporation
• Original Assignee: NVIDIA Corporation
• Inventors: CHEN, Sien, XU, Shuang
• Granted Patent: US 10,536,709 B2
• Time in Patent Office: Days
• US Class Current: 382/154
• CPC Class Codes: H04N 19/29 involving scalability at the object level, e.g. video object layer [VOL]
• CPC Class Codes: H04N 19/597 specially adapted for multi-view video sequence encoding

Abstract

In one embodiment, a method of prioritized compression for 3D video wireless display, the method comprising: inputting video data; abstracting scene depth of the video data; estimating foreground and background for each image of the video data; performing different kinds of compressions to the foreground and background in each image; and outputting the processed video data. Thus, the image quality is not affected by the data loss during the wireless transmission.

Claims

1. A method comprising:

inputting three-dimensional (3D) video data;

abstracting scene depth of the 3D video data;

estimating foreground and background for each image of the 3D video data;

performing different kinds of compressions to the foreground and background in each image; and

outputting the processed video data.

2. The method of claim 1, wherein the inputting 3D video data comprises DMA (direct memory access) and FIFO (first in first out).

3. The method of claim 1, wherein the abstracting scene depth of the 3D video data comprises:

differentiating the type of the 3D video data; and

if the type of the 3D video data is gaming video, copying scene depth from a Z-Buffer.

4. The method of claim 1, wherein the abstracting scene depth of the 3D video data comprises:

differentiating the type of the 3D video data; and

if the type of the 3D video data is stereo stream, estimating motion between a left image and a right image.

5. The method of claim 1, wherein the estimating foreground and background comprises searching a depth threshold in a quantity-depth histogram of each frame.

6. The method of claim 1, wherein the performing different kinds of compressions to the foreground and background in each image comprises:

performing high profile stereo coding to the foreground; and

performing base profile mono coding to the background.

7. The method of claim 1, wherein the performing different kinds of compressions of the foreground and background in each image comprises the compression ratio of the foreground is less than that of the background.

8. The method of claim 7, wherein the compression ratio of the foreground is zero and the compression ratio of the background is 85%.

9. The method of claim 1, wherein the performing different kinds of compressions to the foreground and background in each image comprises the compression rate of the foreground is less than the compression rate of the background.

10. The method of claim 1, wherein the performing different kinds of compressions to the foreground and background in each image comprises the compression rate of the foreground is 80 Mbps (megabits per second) and the compression rate of the background is 20 Mbps.

11. A system comprising:

a data inputting unit for inputting three-dimensional (3D) video data;

a data processing unit for abstracting scene depth of the 3D video data, estimating foreground and background for each image of the 3D video data, and performing different kinds of compressions to the foreground and background in each image; and

a data outputting unit for outputting the processed video data,

the data processing unit coupled to the data inputting unit and the data outputting unit.

12. The system of claim 11, wherein the abstracting scene depth of the 3D video data comprises:

differentiating the type of the 3D video data; and

if the type of the 3D video is gaming video, copying the scene depth from a Z-Buffer.

13. The system of claim 11, wherein the abstracting scene depth of the 3D video data comprises:

differentiating the type of the 3D video data; and

if the type of the 3D video data is stereo stream, estimating the motion between a left image and a right image.

14. The system of claim 11, wherein the estimating foreground and background comprises searching a depth threshold in a quantity-depth histogram of each frame.

15. The system of claim 11, wherein the performing different kinds of compressions to the foreground and background in each image comprises:

performing high profile stereo coding to the foreground; and

performing base profile mono coding to the background.

16. The system of claim 11 wherein the performing different kinds of compressions of the foreground and background in each image comprises the compression ratio of the foreground is less than that of the background.

17. A computer program resident on a computer readable medium, operable for executing instructions to perform a method comprising:

inputting three-dimensional (3D) video data;

abstracting scene depth of the 3D video data;

estimating foreground and background for each image of the 3D video data;

performing different kinds of compressions to the foreground and background in each image; and

outputting the processed video data.

18. The computer program of claim 17, wherein the estimating foreground and background comprises searching a depth threshold in a quantity-depth histogram of each frame.

19. The computer program of claim 17, wherein the abstracting scene depth of the 3D video data comprises:

differentiating the type of the 3D video data; and

if the type of the 3D video data is gaming video, copying scene depth from a Z-Buffer.

20. The computer program of claim 17, wherein the performing different kinds of compressions of the foreground and background in each image comprises the compression ratio of the foreground is less than that of the background.