边缘相似度及其在散斑噪声抑制算法比较中的应用

Edge Similar Degree and Its Application in the Comparison of Suppressing Speckle Noise Algorithm

对于滤波算法,噪声抑制能力和边缘保持能力一直是考核其性能的两个重要指标。对于前者,通常都有相应的参数进行表征,而对于后者,通常却只能依靠人眼进行主观评判。根据标准边缘图像和待测边缘图像边缘点位置的相似性,定义了边缘相似度参数,用来表征滤波算法的边缘保持能力。将此参数应用于高斯噪声图像和椒盐噪声图像滤波算法的比较中,取得了预期的效果:此参数值准确地反映了各滤波算法的边缘保持能力。对激光雷达图像而言,散斑噪声滤波算法的边缘保持能力具有非常重要的意义,为此,将边缘相似度用于散斑噪声滤波算法的边缘保持能力比较,结果表明Lee滤波算法和增强Lee滤波算法具有较强的边缘保持能力,而常规的均值滤波和中值滤波边缘保持能力较差。

Suppression ability and edge=preserving ability are two important indexes to evaluate a filter algorithm. Proper parameter is used to measure the former, but usually only subjective judgment by human eyes can be used in evaluating the latter one. Edge similar degree parameter was defined to measure the edge^preserving ability of a filter algorithm in the paper based on the similarity between the position of edge points in standard edge image and the edge image to be measured. The parameter was used to compare the filter algorithms in processing Gaussian noise affected images and salt ＆ pepper noise affected images. Prospective result was gained and approved that the parameter value measured exactly the edge preserving ability of filter algorithms. The edge-preserving ability of speckle noise filter algorithm is very important to laser radar image. So, the edge similar degree was used to measure the edge-preserving ability of speckle noise filter algorithms. The result showed that Lee filter algorithm and enhanced Lee filter algorithm have preferable edge-preserving ability. The edge-preserving ability of average filter and median filter is dissatisfactory.