基于静态小波变换改进邻域系数法的遥感图像滤波  

Remote Sensing Image Filtering Using Improved NeighShrink Based on Discrete Stationary Wavelet Transformation

作　　者：武海洋[1,2] 王慧[1] 樊菊[2] 

机构地区：[1]解放军信息工程大学测绘学院 [2]68011部队

出　　处：《海洋测绘》2011年第2期24-27,共4页Hydrographic Surveying and Charting

基　　金：国家自然科学基金项目(60778051)

摘　　要：针对NeighShrink滤波算法存在的不足,提出了改进的阈值估计策略和系数收缩方案,并将改进算法与静态小波变换相结合进行图像滤波,取到了较好的滤波效果。实验验证了改进算法的有效性,以峰值信噪比PSNR作为客观评价指标,通过与经典方法的比较,对该算法的滤波性能做出了客观评价。Filtering is one of the essential tasks in the processing and applications of remote sensing images,and image filtering based on wavelet transformation has now become the mainstream in the area.Aimed at the disadvantages in NeighShrink,the author proposed an improved strategy for threshold estimation and coefficients shrinkage,and the improved method based on DSWT was implemented.Experimentation validated the algorithm,and the PSNR was adopted to evaluate the performance of improved algorithms.

关 键 词：静态小波变换 图像滤波 邻域小波系数 阈值 峰值信噪比 

分 类 号：P237[天文地球—摄影测量与遥感] TP751[天文地球—测绘科学与技术]