Multi-iteration wavelet zero-tree coding for image compression

Here, we present a modification of Shapiro's embedded zerotree wavelet algo rithm (EZW) for image codec. Shapiro's technique is based on the wavelet tr ansform and on the self-similarity inherent in images. In the EZW, the wave let transform (WT) coefficients, which provide a multiresolution representa tion of the image, are arranged according to their significance across scal es using a small symbol set (zerotree (ZT) coding). An analysis of the symb ol entropy shows that better compression rates can be obtained when two or more iterations of the original algorithm are combined. Consequently, we pr oposed a modification of Shapiro's original algorithm which we called multi -iteration EZW designed ro optimise the combination of ZT and Huffman codin g. We studied the behaviour of the multi-iteration algorithm in terms of im age quality and bit-rate for natural and medical images. Our findings show that for a given image quality the multi-iteration algorithms and particula rly the two-iteration EZW produce lower bit-rates than Shapiro's. In additi on, we suggest that the idea of multi-iteration can be generalised to other techniques based on ZT coding. (C) 2000 Elsevier Science B.V. All rights r eserved.