Abstract: This paper proposes the adaptive subband decomposition algorithm and context-based entropy arithmetic coding. The adaptive subband decomposition algorithm changes the statistics of transformed coefficients and improves the coding performance in Peak Signal-to-Noise Ratio (PSNR) in low bit rate. In addition, the context model exploits the dependence between subbands and between adjacent pixels and elevates the compression ratio. Experimental results show that the proposed algorithm has the better compression performance than SPIHT and at the time preserves the property of embedded zero tree coding.