The Compression Processing Of Radar Image Based On The Compressed Sensing Theory

Compressed sensing is a very popular research topic at home and abroad in recent years, and it caused widespread attention in many ways of the communication disciplines. For radar, in order to have better range resolution, the signal bandwidth tend to be larger, and the large signal bandwidth is bound to make radar system to face some problems such as storage and handling large amount of data.Compressed sensing theory satisfies the data compression needs of radar exactly.This paper did the research about radar image’s compression and reconstruction based on the compressed sensing theory.The concrete work as following.1)Firstly, discussed the basic theory of compressed sensing and some involved mathematical basis in detail. Analyzed the three elements of compressed sensing theory: the sparse matrix, the measurement matrix and the signal reconstruction algorithm emphatically.2)Then,did some simulation experiment when the compressed sensing theory applied in the concrete compression and reconstruction of radar image.Determined the most suitable sparse matrix (Discrete Wavelets), the most suitable observing matrix(Gauss random matrix)and the most suitable signal reconstruction algorithm(Orthogonal matching pursuit) by the simulation experiment.3)Finally,using C language to realize the method based on compressed sensing theory. The program covers the whole radar image sparse, compression and reconstruction process. The sparse matrix, measurement matrix and the reconstruction algorithm are all filtered out from the simulation experiments. Moreover, in order to achieve the engineering application, this paper designed a system of radar image’s compression, reconstruction and displaying based on DSP, and verificated some key technologies by F2812 development board, this system includes the data transmission module, image’s compressing&reconstruction module and the image displaying module. Besides, on the data transmission module, designed a set of data "code", "decode"rules according to the characteristics of radar image data and serial transmission.The experimental results show that using the compressed sensing theory to reconstruction radar image has the advantages of high precision and fast speed when chosen the right sparse matrix, measurement matrix and reconstruction algorithm. So,the compressed sensing theory applied in radar image’s compression and reconstruction is feasible and has great potential.