Study Of Image Compression Based On Vector Quantization

As an efficient data compression technique, vector quantizaton is known for its low bit rate, simple decoding and minor distortion. It has been widely applied to image and speech compression. The paper deals with image compression based on vector quantization, detailedly expounds its elemental principle, relative conception,and present development, deeply explores its two key technique--codebookgeneration and codeword searching, summarizes and analyzes present typical algorithms, and presents modified algorithms.The paper analyzes binary-split gradient & threshold initial codebook generation -algorithms, codebook generation algorithms based on Kohonen self-organizing feature map neural network, a fast codeword searching algorithm using L2-norm pyramid data structure, side-match vector quantization algorithms, and a fuzzy classified vector quantization algorithm, systematicly explores their application to image compression, computer simulation results show that they are practical and efficient.The paper presents a modified fuzzy Kohonen neural network clustering algorithm fit for image compression, which is applied to codebook generation, theoretic analysis and computer simulation results shows that it can solve some problems which exist in ordinary Kohonen self-organizing feature map neural network algorithms and improve encoding quality.The paper also presents a fast correlation-side match algorithm, which is applied to codeword searching for image vector quantization and has a good result.