Image Compression Using Wavelet Transform and Vector Quantization

Abstract

In this paper a modified LBG vector quantization alglrithm id presented and used with wavelet transform algorithm to compress the image. The modified LBG algorithm provides good representation for training vectors carrying important information which may repesent rare vectors in the training vectors. So this algorithm overcomes the problem of the traditional LBG algorithm that provides good representation for training
vectors which occur frequently while rare vectors are not well represented. The use of the modified LBG algorithm with wavelet transform give better reconstructed image quality than the traditional LBG algorithm because the rare vectors hich carry important information are better represented. The two algorithm give the same compression ratio.