Research On SAR Image Denoising Algorithm Based On NSCT

Synthetic aperture radar(SAR)system is an important research direction in the field of microwave remote sensing.It can improve radar information acquisition ability depending on its characteristics of all-time,all-weather,long distance and high resolution.Therefore it is widely used in the field of agriculture,oceans and military field.But,SAR is a coherent imaging system,which makes it produce speckle noise inevitably in the process of imaging.When speckle noise increases,SAR image will become fuzzy,even lead to the missing of characteristics target.Speckle denoising is the first step of SAR imaging post-processing,and the noise suppression method has been a hot issue in the field of SAR signal processing.The main work of the paper is to study speckle noise denoising algorithms in SAR image.In view of the deficiencies in fourth-order PDE model,a new algorithm of SAR image denoising is proposed based on NSCT.The main research content is summarized as follows:(1)The imaging mechanism of SAR image is studied,and the multiplicative model and statistical properties of speckle noise are analyzed.Classical coherence suppression algorithm and anisotropic diffusion model used in SAR image are introduced.Known through simulation experiment,the best diffusion model is fourth-order PDE model,which has good speckle denoising effect.(2)An improved fourth-order PDE denoising algorithm is proposed based on NSCT.First of al.the SAR image is decomposed into multi-scale and multi-direction sub-band images.Afterwards,adaptively select gradient threshold according to the coefficients of each sub-band image.Through the simulation experiments,it could be found that the improved algorithm has a good ability for preserving,which is based on the self-characteristics of each sub-band image.(3)Due to the facts that when fourth-order PDE is used to filter,the flat areas of the image will produce "spots",for this shortage,a new denoising algorithm based on fourth-order PDE and butterworth low-pass filter is proposed.In this algorithm,the image is decomposed into low frequency coefficient and high frequency coefficients based on NSCT.Different algorithms are adopted to deal with different areas.Simulation experiments prove the validity of the proposed algorithm.