SAR Image Scalable Compression Based on the Modification of EZBC Algorithm Using Wavelet Packet Transform

In this paper, a block-based method in wavelet domain for SAR image coding is presented based on the EZBC algorithm. To better preserve edges and textural structure in SAR images, instead of using traditional dyadic wavelet transform, a packet wavelet is employed. To improve the efficiency of the EZBC algorithm for SAR image coding, a modified version of this algorithm, is proposed. This modified version, which is called MEZBC, is better adapted to wavelet packet coefficients. The proposed algorithm is fully SNR scalable and supports block-based coding of wavelet coefficients. MEZBC is also further modified to support regions of interest (ROI) coding in SAR images. experimental results show that the proposed algorithm provides a better performance for SAR image coding than other well-known algorithms in almost all bit rates.