Sar Raw Data Compression with Dct-baq

In this study Synthetic Aperture Radar (SAR) raw data compression techniques are introduced. Two encoding approaches are generally used. Direct encoding in spatial domain and transform encoding in transform domain. A well-known spatial domain compression technique is based on Block Adaptive Quantization (BAQ)[3,6]. BAQ encoding is implemented by dividing the SAR Raw data set into data blocks of small size and by estimating for each block, the standard deviation value. In transform domain we use discrete cosine transform (DCT). SAR Raw data transform coefficients have high dynamic range. This causes reduction for compression performans. We have applied new approach for reduction of dynamic range. We have coded DC component of the transform coefficients with 8 Bits for each data block and transmit standard deviation and DC component seperately from transform domain coefficients. Dynamic range reduction of transform domain coefficients improves compression performans results versus BAQ