Vector Quantization Method for Image Compression Based on GA and LBG Clustering Algorithm

Vector quantization is a very important image compression algorithm and its key is the design of codebook.The classic LBG clustering algorithm results in different performance because of sensitive selection of the initial cluster center.This paper applied LBG clustering algorithm into genetic algorithm to optimize cluster center,which uses the advantages of high local search ability of LBG and global optimization ability of GA.The hybrid method can not only improve the quality of codebook but also speed algorithm convergence.