System And Method For Graphics Compression

• US 20160196632A1
• Filed: 01/02/2015
• Published: 07/07/2016
• Est. Priority Date: 01/02/2015
• Status: Active Grant

• Current Assignee: Avago Technologies General IP PTE Limited (Broadcom, Inc.)
• Original Assignee: Broadcom Corporation (Broadcom, Inc.)
• Inventors: Vadakkumthodam Radhakrishnan, Praveen, Schoner, Brian Francis, Pande, Anand
• Granted Patent: US 10,462,465 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G06T 9/00 Image coding bandwidth or redundancy reduction for static pictures H04N1/41; coding or decoding of static colour picture signals H04N1/64; methods or arrangements for coding, decoding, compressing or decompressing digital video signals H04N19/00
• CPC Class Codes: H04N 19/136 Incoming video signal characteristics or properties
• CPC Class Codes: H04N 19/176 the region being a block, e.g. a macroblock
• CPC Class Codes: H04N 19/463 by compressing encoding parameters before transmission
• CPC Class Codes: H04N 19/593 involving spatial prediction techniques

Abstract

A system and method for processing graphics are provided. Pixel data may be received for a pixel block. Endpoints for the values of the pixels in the pixel block may be determined. A weight for the pixels in the pixel block may be determined in four dimensions corresponding to the endpoints. A compressed data block representative of the pixel block may be generated in response to the endpoints for the pixel block and the weight for the pixels of the pixel block in the four dimensions corresponding to the endpoints.

Claims

1. A graphics processor comprising:

endpoint selection circuitry operable to receive pixel data for a pixel block and determine endpoints for values of the pixels in the pixel block;

weight selection circuitry operable to receive pixel data for the pixel block, the weight selection circuitry operable to determine a weight for each pixel of the pixel block in each of four dimensions corresponding to the endpoints selected for the block by the endpoint selection circuitry; and

packing circuitry operable receive the endpoints for the pixel block and the weight for each pixel of the pixel block in each of four dimensions corresponding to the endpoints, the packing circuitry being operable to generate a compressed data block representative of the pixel block.

2. The graphics processor according to claim 1, wherein one of the four dimensions is an independent alpha dimension.

3. The graphics processor according to claim 2, wherein the four dimensions include a red dimension, a green dimension, and a blue dimension.

4. The graphics processor according to claim 1, wherein the pixel block is a non-square pixel block and the pixel block includes a greater number of pixels along a raster direction than a non-raster direction.

5. The graphics processor according to claim 4, wherein the pixel block includes a width of a single pixel along the non-raster direction.

6. The graphics processor according to claim 5, wherein the pixel block includes a width of eight pixels along a raster direction and a width of two pixels along a non-raster direction.

7. The graphics processor according to claim 5, wherein the pixel block includes a width of sixteen pixels along a raster direction and a width of one pixel along a non-raster direction.

8. A method for processing graphics comprising:

receiving, in a graphics processor, pixel data for a pixel block;

determining, in the graphics processor, endpoints for values of the pixels in the pixel block;

determining, in the graphics processor, a weight for each pixel of the pixel block in each of four dimensions corresponding to the endpoints selected for the block; and

generating, in the graphics processor, a compressed data block representative of the pixel block in response to the endpoints for the pixel block and the weight for each pixel of the pixel block in each of the four dimensions corresponding to the endpoints.

9. The method according to claim 8, further comprising:

determining, in the graphics processor, a weight for each pixel of the pixel block in each of three dimensions corresponding to the endpoints selected for the block, wherein at least one of the three dimensions is different than each of the four dimensions; and

generating, in the graphics processor, a compressed data block representative of the pixel block in response to the endpoints for the pixel block and the weight for each pixel of the pixel block in each of three dimensions corresponding to the endpoints.

10. The method according to claim 9, wherein the determining a weight for each pixel of the pixel block in each of the three dimensions and the generating a compressed data block in response to the endpoints for the pixel block and the weight for each pixel of the pixel block in each of the three dimensions occur in parallel with the determining a weight for each pixel of the pixel block in each of the four dimensions and the generating a compressed data block in response to the endpoints for the pixel block and the weight for each pixel of the pixel block in each of the four dimensions.

11. The method according to claim 9, wherein the determining the weight for each pixel of the pixel block in each of the three dimensions and the determining the weight for each pixel of the pixel block in each of the four dimensions occur simultaneously.

12. The method according to claim 9, wherein the generating the compressed data block in response to the endpoints for the pixel block and the weight for each pixel of the pixel block in each of the three dimensions and the generating the compressed data block in response to the endpoints for the pixel block and the weight for each pixel of the pixel block in each of the four dimensions occur simultaneously.

13. The method according to claim 8, wherein selecting endpoints further comprises:

determining whether values of the pixels in the pixel block exceed a value threshold;

determining whether a number of pixels, that exceed the value threshold, exceed a pixel count threshold; and

setting the endpoints to a default value in response to the number of pixels exceeding the pixel count threshold.

14. The method according to claim 13, wherein the default value for the endpoints include a maximum range value and a minimum range value.

15. The method according to claim 13, wherein selecting endpoints further comprises:

expanding a number of bits allocated to the weight for each pixel of the pixel block in response to setting the endpoints to the default value.

16. The method according to claim 8 wherein the determining the weight for each pixel of the pixel block includes a division.

17. The method according to claim 16, wherein a dividend of the division comprises a value of the pixel and a divisor of the division comprises a range of the endpoints.

18. The method according to claim 17, wherein a remainder of the division is used as an estimate of an error for the weighting.

19. A method for processing graphics comprising:

receiving, in a graphics processor, pixel data for a pixel block;

determining, in the graphics processor, first endpoints for first component values of the pixels in the pixel block, the endpoints corresponding to a first component of a multi-component plane;

determining a first error for the first endpoints;

determining, in the graphics processor, second endpoints for second component values of the pixels in the pixel block, the second endpoints corresponding to a second component of the multi-component plane;

determining a second error for the second endpoints;

comparing the first error for the first endpoints to the second error for the second endpoints;

selecting one of the first endpoints and the second endpoints as selected endpoints in response to the comparing of the first error to the second error; and

generating, in the graphics processor, a compressed data block representative of the pixel block in response to the selected endpoints for the pixel block and a weight for each pixel of the pixel block corresponding to the selected endpoints.

20. The method according to claim 19, wherein a division operation is performed for each pixel of the pixel block, a dividend of the division operation comprises a value of the pixel and a divisor of the division operation comprises a range of the endpoints, a remainder of the division operation is used as an estimate of the error for weighting.