Conference Proceedings

Abstract

The EMBDDED ZEROTREE WAVELET (EZW) algorithm, as presented by J. Shapiro, is a simple yet powerful algorithm, in which bit-streams are generated in order of their significance in containing the image information [1]. The original EZW algorithm scans the entire wavelet decomposed image, at a stroke, during each pass. This work presents a modified method for coding images using EZW method, which works on the principal of fragmentation. The proposed method takes the smallest unit cell, generated from the wavelet decomposed image, to encode at a time .This makes the encoding times independent of the level of wavelet decomposition. Results show that the proposed algorithm is more efficient in performance, in terms of encoding times, as compared to the original algorithm. Moreover, the difference between the encoding times using the original and proposed method, tends to increase with an increase in the dimensions of the image under test.