Embedded Zerotree wavelet coding of image sequence

Abstract

In this paper we present an image sequence coding system based on Embedded Zerotree Wavelet algorithm (EZW). Difference between the image in the coder and the reconstructed previous image in the decoder is used as technique for removing the temporal redundancies. The first image is encoded in intra-mode by EZW algorithm and a specific binary codebook CB1. The subsequent images in the sequence are encoded by performing the difference between the reconstructed previous image in the decoder and the current image in the coder; this difference (residual image) is then encoded by EZW algorithm and a specific binary codebook CB2. Simulations are operated on Claire and Alexis sequences. The results show that the system can provides best reconstruction quality as well objectively as subjectively for a minimum given bit rate. Progressive transmission, rate control for constant bit-rate and rate scalability are the main characteristics of this system.