Wavelets image data compression

The fast development of multimedia computing has led to the demand of using digital images. The manipulation, storage and transmission of these images in their raw form is very expensive, it significantly slows the transmission and makes storage costly. Many techniques are now available and much effort is being expended in determining the optimum compression technique. Recently, compression techniques using wavelet transforms (WTs) have received great attention, because of their promising compression ratio, flexibility in representing images and its ability to take into account the human visual system. In this paper, the authors combine a wavelet transform with vector quantization, using a modified version of the LBG algorithm using the partial search partial distortion (PSPD) scheme, for coding the wavelet coefficients to speed up the scheme in codebook generation and the search required for nearest neighbour codevector of input image. The wavelet transform is used to obtain a set of different frequency subbands of the image; the image is decomposed using a pyramidal algorithm architecture. According to Shannon´s rate distortion theory, the wavelet coefficients are vector quantized using multiresolution codebooks. The proposed scheme can save 70-80% of the vector quantization (VQ) encoding time compared to fully search VQ and reduced arithmetic complexity without sacrificing performance