Methods for compressing and decompressing texture tiles and apparatuses using the same

Patent metadata

Field	Value
Publication number	US-10074040-B2
Application number	US-201615364998-A
Country	US
Kind code	B2
Filing date	Nov 30, 2016
Priority date	Nov 18, 2016
Publication date	Sep 11, 2018
Grant date	Sep 11, 2018

Abstract

The invention introduces a method for compressing texture tiles, which contains at least the following steps: classifying each pixel of a texture tile into one of a plurality of groups and generating a bitmap of the texture tile, wherein the bitmap contains information indicating to which group each pixel of the texture tile belongs; reducing dependencies between pixels of each group; lossless-encoding each pixel of the texture tile to generate a first compression result; and outputting the first compression result and the bitmap.

First claim

What is claimed is:

1. A method for compressing texture tiles, comprising:
classifying each pixel of a texture tile into one of a plurality of groups and generating a bitmap of the texture tile, wherein the bitmap comprises information indicating to which group each pixel of the texture tile belongs;
reducing dependencies between pixels of each group;
lossless-encoding each pixel of the texture tile to generate a first compression result;
outputting the first compression result and the bitmap;
lossless-compressing each pixel of the texture tile to generate a second compression result;
determining whether a length of the second compression result is longer than a target length; and
when the length of the second compression result is longer than the target length, generating the bitmap of the texture tile.

2. The method of claim 1 , wherein the groups comprise a foreground and a background.

3. The method of claim 2 , wherein the step of classifying each pixel of a texture tile into one of a plurality of groups and generating a bitmap of the texture tile comprises:
generating an R-channel histogram, a G-channel histogram and a B-channel histogram of the texture tile and calculating an R-channel boundary value, a G-channel boundary value and a B-channel boundary value accordingly;
determining whether an R-value, a G-value and a B-value of each pixel are greater than the R-channel boundary value, the G-channel boundary value and the B-channel boundary value, respectively; and
when two or more of the R-value, the G-value and the B-value of each pixel are greater than the respective boundary values, labeling this pixel as a foreground pixel; otherwise, labeling this pixel as a background pixel.

4. The method of claim 1 , wherein the groups comprise a foreground, an intermediary and a background.

5. The method of claim 4 , wherein the step of classifying each pixel of a texture tile into one of a plurality of groups and generating a bitmap of the texture tile comprises:
generating an R-channel histogram, a G-channel histogram and a B-channel histogram of the texture tile and calculating an R-channel middle range, a G-channel middle range and a B-channel middle range accordingly;
determining whether an R-value, a G-value and a B-value of each pixel are greater than an upper value of the R-channel middle range, an upper value of the G-channel middle range and an upper value of the B-channel middle range, respectively;
determining whether the R-value, the G-value and the B-value of each pixel are less than a lower value of the R-channel middle range, a lower value of the G-channel middle range and a lower value of the B-channel middle range, respectively; and
when two or more of the R-value, the G-value and the B-value of each pixel are greater than the respective upper values, labeling this pixel as a foreground pixel; when two or more of the R-value, the G-value and the B-value of each pixel are less than the respective lower values, labeling this pixel as a background pixel; otherwise, labeling this pixel as an intermediary pixel.

6. The method of claim 1 , wherein the step of reducing dependencies between pixels of each group comprises:
except for an upper-left pixel of each group, an R-value, a G-value and a B-value of each pixel of this group are subtracted from an R-value, a G-value and a B-value of the upper-left pixel of this group, respectively.

7. An apparatus for compressing texture tiles, comprising:
a bitmap generator;
a compressor; and
a data generation controller, coupled to the bitmap generator and the compressor, directing the bitmap generator to classify each pixel of a texture tile into one of a plurality of groups and generate a bitmap of the texture tile, wherein the bitmap comprises information indicating to which group each pixel of the texture tile belongs; and directing the compressor to reduce dependencies between pixels of each group, lossless-encode each pixel of the texture tile to generate a first compression result, and output the first compression result and the bitmap;
wherein the data generation controller directs the compressor to lossless-compress each pixel of the texture tile to generate a second compression result determines whether the length of the second compression result is longer than a target length; and when the length of the second compression result is longer than the target length, directs the bitmap generator to generate the bitmap of the texture tile.

8. The apparatus of claim 7 , wherein the groups comprise a foreground and a background.

9. The apparatus of claim 8 , wherein the bitmap generator generates an R-channel histogram, a G-channel histogram and a B-channel histogram of the texture tile and calculates an R-channel boundary value, a G-channel boundary value and a B-channel boundary value accordingly; determines whether an R-value, a G-value and a B-value of each pixel are greater than the R-channel boundary value, the G-channel boundary value and the B-channel boundary value, respectively; when two or more of the R-value, the G-value and the B-value of each pixel are greater than the respective boundary values, labels this pixel as a foreground pixel; otherwise, labels this pixel as a background pixel.

10. The apparatus of claim 7 , wherein the groups comprise a foreground, an intermediary and a background.

11. The apparatus of claim 10 , wherein the bitmap generator generates an R-channel histogram, a G-channel histogram and a B-channel histogram of the texture tile and calculates an R-channel middle range, a G-channel middle range and a B-channel middle range accordingly; determines whether an R-value, a G-value and a B-value of each pixel are greater than an upper value of the R-channel middle range, an upper value of the G-channel middle range and an upper value of the B-channel middle range, respectively; determines whether the R-value, the G-value and the B-value of each pixel are less than a lower value of the R-channel middle range, a lower value of the G-channel middle range and a lower value of the B-channel middle range, respectively; and when two or more of the R-value, the G-value and the B-value of each pixel are greater than the respective upper values, labels this pixel as a foreground pixel; when two or more of the R-value, the G-value and the B-value of each pixel are less than the respective lower values, labels this pixel as a background pixel; otherwise, labels this pixel as an intermediary pixel.

12. The apparatus of claim 7 , wherein the compressor, except for an upper-left pixel of each group, subtracts an R-value, a G-value and a B-value of each pixel of this group from an R-value, a G-value and a B-value of the upper-left pixel of this group, respectively.

13. A method for decompressing texture tiles, comprising:
receiving a compression result and a text flag;
determining which one of a first algorithm and a second algorithm is used to reduce dependencies between pixels of a texture tile by an encoder;
when the first algorithm is used to reduce dependencies between pixels of the texture tile by the encoder, lossless-decoding the compression result and using a third algorithm associated with the first algorithm to restore dependencies between pixels of the texture tile;
when the second algorithm is used to reduce dependencies between pixels of the texture tile by the encoder, obtaining a bitmap, lossless-decoding the compression result and using a fourth algorithm associated with the second algorithm to restore dependencies between pixels of the texture tile according to the bitmap to generate a first restored result;
receiving a lossy-level flag;
determining wh

Assignees

• Via Alliance Semiconductor Co Ltd

Inventors

• Shen Yemao
• Gu Deming
• Que Heng
• Zhang Wei

Classifications

• H04N19/91

  Entropy coding, e.g. variable length coding [VLC] or arithmetic coding · CPC title

• H04N19/50

  using predictive coding (H04N19/61 takes precedence) · CPC title

• H04N19/46

  Embedding additional information in the video signal during the compression process (H04N19/517, H04N19/68, H04N19/70 take precedence) · CPC title

• H04N19/436

  using parallelised computational arrangements · CPC title

• H04N19/186

  the unit being a colour or a chrominance component · CPC title