基于cs的sar目标识别-智能信息处理专业论文.docxVIP

Abstract Abstract III Abstract Synthetic aperture radar(Synthetic Aperture Radar,referred to aS SAR)is a new kind of microwave imaging radar．It is capable of penetrating soil and vegetation．SAR image target recognition plays a more and more important role in the national economy and national defense construction because of its a11一weather，wide，strong penetrable ability and supplying detailed ground mapping material and images in atrocious weather with hi曲 resolution and．So how to interpretate SAP,image arouse wide attention．There is a lot of speckle noise in SAR image．The classical algorithms generally take complex pre-processing before identification，and the pre-processing has great influence on the performance of recognition．The complexity of pre-processing makes it hard to build a very efficient target recognition system for SAR images． Compressed sensing theory(CS)shows that if the signal is sparse，it can be precisely reconstructed诵也few observations．When CS is applyed to target recognition， it can achieve a hi曲performance with very few observations．The recognition based on compressed sensing algorithm provides a new insight into crucial issues in recognition： good robustness to occlusion．So the recognition via compressed sensing is very suitable for SAR target recognition．In this paper，the algorithm is applied to target recognition of SAR image without image pre-processing．This thesis studies SAR image target recognition via compressed sensing，and analysis the feasibility．To address the above issues and the characteristics of SAR image target recognition，we propose a new method to recognition and achieve good results．This research is significant in the development of a new SAR recognition system without reconstructing spatial images and traditional pre-processing． Experiments results show that 011Imethod Call effectively identify the target without image pre-processing．Particularly,with down-sampling method for feature extraction，it still achieved good results，