Applications Of Wavelet Transform And Vector Quantization Technique In Image Compression

With the fast developing of network and multimedia application, people pay more and more attentions to the technology of image compression. At present, the technology of image compression based on the wavelet transform is replacing the technology based on DCT or other coding techniques step by step, and becoming the preferred method in the new international image compression standards. The applications of wavelet transform in image compression are explored.Firstly, a new algorithm of vector quantization based on the progressive constructive clustering and self-organizing feature map is proposed. The experimental results show that the reconstruction quality of this algorithm has an advantage of 0.2～1.5dB over the former two methods at the same bit rate.Secondly, mean energies of high-frequency subbands and mean ranges of three sub low-frequency subbands of image transformed by wavelet are studied, then proposed a simple rule by which the image can be divided into two different areas named important direction area and unimportant direction. A multi-scheme algorithm of vector quantization in image compression based on wavelet transform is proposed, and experimental results reveal the scheme performance well.Finally, a scheme of region-of-interest image coding is proposed based on orientation tree and vector classification. The simulation results show that this method can still hold fairly good subjective quality at a very low bit rate.