US 11,776,164 B2
Content-adaptive tiling solution via image similarity for efficient image compression
Ajay Bedi, Hamirpur (IN); and Rishu Aggarwal, Delhi (IN)
Assigned to Adobe, Inc., San Jose, CA (US)
Filed by Adobe, Inc., San Jose, CA (US)
Filed on Dec. 22, 2022, as Appl. No. 18/145,118.
Application 18/145,118 is a continuation of application No. 17/120,175, filed on Dec. 13, 2020, granted, now 11,562,508.
Prior Publication US 2023/0126890 A1, Apr. 27, 2023
This patent is subject to a terminal disclaimer.
Int. Cl. G06T 9/00 (2006.01); H03M 7/30 (2006.01); G06T 3/40 (2006.01)
CPC G06T 9/001 (2013.01) [G06T 3/40 (2013.01); H03M 7/3059 (2013.01); H03M 7/6011 (2013.01)] 20 Claims
OG exemplary drawing
 
1. A method for efficiently storing a set of image files, the method comprising:
generating a first plurality of tiles for a first image, wherein tiles of the first plurality of tiles include pixels that are within a threshold color intensity of one another;
generating a second plurality of tiles for a second image, wherein tiles of the second plurality of tiles include pixels that are within the threshold color intensity of one another;
generating a first tile cluster that includes at least one tile from the first image and at least one tile from the second image;
removing a duplicate tile from the first tile cluster, wherein pixels of the duplicate tile share a same color intensity as another tile in the first tile cluster, and wherein remaining tiles after removing the duplicate tile constitute a filtered first tile cluster;
generating a second tile cluster that includes tiles from the filtered first tile cluster that share a common encoding parameter; and
compressing one or more tiles of the second tile cluster, thereby providing one or more compressed tiles.