Performance evaluation of hybrid coding of images using wavelet transform and predictive coding

Abstract

Image compression techniques are necessary for the storage of huge amounts of digital images using reasonable amounts of space, and for their transmission with limited bandwidth. Several techniques such as predictive coding, transform coding, subband coding, wavelet coding, and vector quantization have been used in image coding. While each technique has some advantages, most practical systems use hybrid techniques which incorporate more than one scheme. They combine the advantages of the individual schemes and enhance the coding effectiveness. This paper proposes and evaluates a hybrid coding scheme for images using wavelet transforms and predictive coding. The performance evaluation is done using a variety of different parameters such as kinds of wavelets, decomposition levels, types of quantizers, predictor coefficients, and quantization levels. The results of evaluation are presented.