SAR RAW DATA PROCESSING APPROACH BASED ON A COMBINATION OF LBG ALGORITHM AND COMPRESSED SENSING

Aimed at the problem of how to diminish SAR raw data apparently and realize SAR imaging effectively, a new approach for processing SAR raw data combined with Linde-Buzo-Gray (LBG) algorithm and Compressed Sensing (CS) is proposed in this paper. For SAR returned signals, CS is engaged to reduce the sampling number in the pulse duration, and LBG algorithm as a classical vector quantization (VQ) method, is employed to diminish encode number of every sample value. Next, data reconstruction process still contains the two ordinal steps according to LBG algorithm and CS theory, respectively. On the basis of that, the traditional SAR imaging method, Frequency Scaling (FS) algorithm, is carried out to achieve the final SAR image. Simulation results show that the high quality SAR image can be achieved on condition of the SAR raw data is diminished furthermore obviously, which is compared with the traditional method.