A Design of JPEG Quantization Table using Genetic Algorithms

L.F. Costa and A.C.P. Veiga (Brazil)

Keywords

Genetic Algorithms, Quantization Table, and NaturalSelection.

Abstract

In several situations, images are processed in JPEG standard with a Quantization Table without any knowledge of where does the coefficient come from [1], and there is little literature to produce a robust quantization table. The fidelity loss in JPEG coding occurs entirely in quantization [2], and it becomes very important to the compression/fidelity performance of JPEG coding. This article has the purpose to present a new method of generating a robust Quantization Table (Q Table) using Genetic Algorithms. The development of The Quantization Table using Genetic Algorithms with the JPEG standard is a important method to create a better Quantization Table for each class of images and different sizes by the natural selection, who also give good conditions to change anytime the parameters of the program to produce better results.

Important Links:



Go Back