Open AccessAN ANALYSIS OF IMAGES BEFORE COMPRESSION IN PNG FORMAT
Author(s) -
Alexander Shportko
Publication year - 2010
Publication title -
computing
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.184
H-Index - 11
eISSN - 2312-5381
pISSN - 1727-6209
DOI - 10.47839/ijc.9.2.712
Subject(s) - pixel , computer science , image compression , data compression , compression (physics) , algorithm , lossless compression , entropy (arrow of time) , image (mathematics) , set (abstract data type) , block (permutation group theory) , artificial intelligence , computer vision , image processing , mathematics , programming language , geometry , materials science , physics , composite material , quantum mechanics
An influence of predictor’s different variants and their combination on PNG image compression is examined in this article. Fragments in C programming language for estimation of pixel lines entropy before and after application of context-dependent algorithm have been presented. A method for image analysis before the compression through breakdown of pixel lines with close entropy values into minimum blocks is offered. The method is based on the choice of an effective compression variant from five possible ones for every minimum block using dynamic programming and on the combination of contiguous minimum blocks in the homogeneous blocks of lines and on adjustment of their code parameters. It is experimentally showed that the images analysis, for example, allows to decrease the continuous tone images compression aspect of Kodak True Color PNG format image set on the average up to 12,37 bpp.