Color image coding using wavelet pyramid coders

Despite the existence of the current standard for still image compression namely JPEG (Joint Photographic Experts Group), alternative techniques for image compression are being investigated due to the need for fast and high quality storage and transmission of large datasets for distribution through the Internet or in CD-ROM. The use of the wavelet transform for multi-resolution pyramid decomposition of images for data compression and transmission has already been successfully employed by a number of researchers using the common frame of linear subband filtering techniques. However, to achieve a desired compression ratio for a specific class of images, an optimized compression algorithm needs to be developed by considering all stages, i.e. from initial transform mapping to final encoding. This paper addresses the suitability and requirement of further development of a wavelet pyramid coding in compressing color images, specifically the color Visible Human digital photographic dataset. The distinct advantages of a newly developed adaptive vector quantization (VQ) over commonly used scalar quantization and conventional VQ of the wavelet decomposed error pyramids are also discussed.