Texture compression techniques
View Patent ↗A texture compression method is described. The method comprises splitting an original texture having a plurality of pixels into original blocks of pixels. Then, for each of the original blocks of pixels, a partition is identified that has one or more disjoint subsets of pixels whose union is the original block of pixels. The original block of pixels is further subdivided into one or more subsets according to the identified partition. Finally, each subset is independently compressed to form a compressed texture block.
1. A texture compression method, comprising:
(a) splitting an original texture having a plurality of pixels into original blocks of pixels;
for each of the original blocks of pixels:
(b) identifying a partition from a predefined set of partitions, the partitions having one or more disjoint and variably shaped subsets of pixels whose union is the original block of pixels;
(c) subdividing the original block of pixels into one or more subsets according to the identified partition;
(d) independently compressing each subset to form a compressed texture block; and
(e) decompressing the compressed texture block to obtain an approximation of the original block of pixels;
(f) comparing the original block of pixels to the approximation of the original block of pixels to obtain a quality parameter for the identified partition;
(g) repeating steps (b) through (f) for each partition in the predefined set of partitions to obtain the quality parameter for each partition;
(h) choosing a final partition for the original block of pixels that yields an optimal quality parameter;
(i) subdividing the original block of pixels into one or more final subsets according to the final partition; and
(j) independently compressing each final subset to form a compressed texture block.
2. The method of claim 1 , further comprising:
for each original block of pixels,
evaluating a function for the original block of pixels; identifying a partition based on the evaluated function; and repeating (c) and (d).
3. The method of claim 1 , wherein the final partition is selected for the original block of pixels if the quality parameter for the identified partition meets a pre-selected threshold and the final partition is selected that either meets the pre-selected threshold or is closest to the pre-selected threshold when none of the partitions meet the pre-selected threshold.
4. The method of claim 1 , wherein (b) further comprises identifying the partition from a predefined set of partitions wherein the subsets have a different number of pixels.
5. The method of claim 1 , wherein the predefined set of partitions includes an explicitly defined base set of partitions and additional sets of partitions derived from the base set.
6. The method of claim 5 , wherein each derived partition is obtained from a partition of the base set by performing unions of some of its subsets.
7. The method of claim 1 , wherein (d) further comprises compressing the pixels in each subset using a lossy compression method.
8. The method of claim 1 , wherein (d) comprises compressing each of the subsets using two explicit endpoint parameters and a number of intermediate points defining a ramp in the color space.
9. The method of claim 1 , wherein (d) comprises compressing each of the subsets using a palletization technique.
10. The method of claim 1 , wherein the quality parameter is based on a root-mean-square error for the block.
11. The method of claim 1 , wherein the quality parameter is based on chrominance error parameter for the block.
12. The method of claim 1 , wherein the quality parameter is based on a luminance parameter for the block.
13. The method of claim 1 , further comprising splitting the texture into regular blocks of 8×8 pixels.
14. The method of claim 1 , further comprising splitting the texture into regular blocks of 4×4 pixels.
15. A computer processing system for texture compression, comprising:
a processor configured to (a) split an original texture having a plurality of pixels into original blocks of pixels, wherein for each of the original blocks of pixels, the processor is further configured to:
(b) identify a partition from a predefined set of partitions, the partitions having one or more disjoint and variably shaped subsets of pixels whose union is the original block of pixels;
(c) subdivide the original block of pixels into one or more subsets according to the identified partition; and
(d) independently compress each subset to form a compressed texture block; and
(e) decompress the compressed texture block to obtain an approximation of the original block of pixels;
(f) compare the original block of pixels to the approximation of the original block of pixels to obtain a quality parameter for the identified partition;
(g) repeat steps (b) through (f) for each partition in the predefined set of partitions to obtain the quality parameter for each partition;
(h) choose a final partition for the original block of pixels that yields an optimal quality parameter;
(i) subdivide the original block of pixels into one or more final subsets according to the final partition; and
(j) independently compress each final subset to form a compressed texture block.