摘要
为克服精致极化Lee滤波算法对极化信息保持不够理想的弊端,本文提出一种改进方法。其基本思路是先应用四分量分解算法将像素分成4种散射类型,然后在方向窗口内去除与中心像素散射特性不同的像素,最后基于同质区的局部统计特性进行滤波处理。试验表明,改进后的滤波算法较传统算法有更强的极化信息保持能力。
An improved method for overcoming the deficiency of Lee refined polarized filter in preserving polarized information was proposed in present study. To implement this method, the paper firstly divided all pixels into four scattering categories based on the four-component decomposition algorithm. Then, pixels with different scattering characteristics to the centralized one were removed in edge-aligned window. Finally, speckle filtering was utilized according to the local statistical properties of homogeneous area. Experimental results showed that the proposed method could perform better in preserving polarized information than traditional one.
出处
《测绘科学》
CSCD
北大核心
2011年第5期136-138,共3页
Science of Surveying and Mapping
基金
国家863计划资助项目(2007AA12Z143)
国家自然科学基金资助项目(40201039
40771157)
关键词
极化SAR
相干斑
精制极化Lee滤波
四分量散射模型
polarized Synthetic Aperture Radar
speckle
Lee refined polarized filter
four-component scattering model