Space adaptive quantization for SAR data compression

A space-adaptive and dynamic block quantization compression scheme suited for Synthetic Aperture Radar missions is presented. The algorithm is an extension of the Flexible Dynamic Block Adaptive Quantization (FDBAQ) [1], [2] to the case the quantizers set for received SAR data are selected after an optimization of the Signal-to-Noise ratio with a constrain on the average bit rate. The FDBAQ scheme allows to obtain better performances compared to the Block Adaptive Quantization (BAQ) [3] by dynamically varying the number of quantization bits according to the received signal power. The proposed Space Adaptive Flexible Block Quantization (SA-FBQ), with the aim of further increasing the overall image quality, allows the compression to cope with the regional reflectivity variations by making available more than one quantizer set; the optimization strategy looks for quantizers within a continuous set. The performances of the considered schemes are evaluated, compared and finally validated using real SAR data acquisitions.