Research On Sidelobe Suppression And Coherent Speckle Suppression Algorithm In SAR Image

Synthetic aperture radar uses radar pulse compression technology,which can work normally in all weather and achieve high-resolution and high-quality remote sensing imaging.Because of its unique advantages,synthetic aperture radar is widely used in target recognition,terrain exploration and other fields,and has great development potential.In practice,if the quality of radar image is poor,it will have a serious impact on target detection and subsequent research.The quality of SAR image will be damaged by sidelobe and speckle interference due to the influence of system and imaging environment.Aiming at sidelobe and speckle jamming,this paper studies the sidelobe suppression and speckle suppression algorithms of SAR images.In this paper,the wavelet transform is used to process the image and the improved spatial apodization method is used to suppress the sidelobe.By combining wavelet threshold denoising and spatial domain filtering,a better denoising effect is achieved while protecting the image edge information.The main research contents are as follows:(1)The related algorithms of sidelobe suppression of synthetic aperture radar image are studied,and an improved sidelobe suppression algorithm combining Spatial Variant Apodization and wavelet transform is proposed.The weighted cosine function model in Spatial Variant Apodization is studied.Although the Spatial Variant Apodization method uses the adaptive cosine weighting function to judge the sidelobe suppression,the threepoint model can not suppress the sidelobe in complex cases.RSVA algorithm is a robust five point filtering algorithm with better side lobe suppression effect.In this paper,RSVA algorithm and wavelet transform are combined to suppress the image sidelobe more carefully.Experimental results show that compared with the traditional Spatial Variant Apodization method,this algorithm has better effect on sidelobe suppression of synthetic aperture radar image.(2)The algorithms of speckle suppression for SAR images are studied.A speckle suppression algorithm combining wavelet threshold denoising and spatial domain filtering is studied.The mathematical model of spatial domain correlation algorithm and the process and details of wavelet domain de-noising are studied,and the effects,advantages and disadvantages of each algorithm are analyzed.The 3D block matching algorithm with good denoising effect but high time complexity is studied,and the process of combining wavelet threshold filtering with other filtering methods is analyzed.This paper analyzes the characteristics of the decomposed layers in the wavelet transform,and combines the wavelet threshold denoising and spatial domain filtering algorithm to suppress the image noise.In order to remove the noise in the image and protect the edge information at the same time,a smoother threshold function and a threshold value with attenuation coefficient are used,combined with the improved frequency domain filtering used in low frequency,to complete the overall optimization.The denoising experiments on real images show that the algorithm can achieve better denoising and edge preservation effects compared with traditional algorithms and emerging algorithms.