Research On Speckle Reduction And Image Enhancement Methods In SAR Images

Synthetic Aperture Radar(SAR)is a new high-resolution imaging Radar with strong penetrability and weather immunity.The coherent imaging mechanism of SAR results in speckles,while the large dynamic range of electromagnetic scattering coefficient results in low contrast of SAR images.Therefore,to improve the visibility and quality,SAR image processing is required.In this thesis,two key problems of SAR image speckle suppression and contrast enhancement are studied,and algorithms are proposed.The SAR image processing software is designed and implemented.The main work include:1)The principle of speckles generation is introduced.The statistic model of SAR image speckles and gray scales are summarized.The character of speckles and image contrast are analyzed,which lays a theoretical foundation for the later research.2)Aiming at the problem of insufficient direction capture capacity in SAR image of non-local means(NLM),a NLM algorithm based on variation coefficient is proposed.An adaptive anisotropic window is constructed,and an adaptive filtering parameter is designed.The new algorithm can suppress speckles of the homogenous region while maintaining the texture of the heterogeneous region more accurately.3)Aiming at the problems of low contrast and mismatch of the dynamic range between SAR image and display equipment,an enhancement algorithm was studied.The membership function is constructed,and the optimization criteria of enhancement model was modified.The algorithm realizes dynamic range compression while effectively retaining the texture details of the image.4)A SAR image processing software framework based on QT API is designed and developed,and common SAR image processing algorithms are integrated.The software realizes practical operation of SAR image processing.The above research content has been verified by simulation data,and the results show that the above algorithm can effectively improve SAR image quality from different angles.Through testing,the software designed and implemented in this paper can meet the enhancement requirements of most SAR images.