论文部分内容阅读
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。
For the filtering algorithm, noise suppression and edge retention has always been to assess the performance of the two important indicators. For the former, there are usually corresponding parameters to characterize, but for the latter, usually only rely on the human eye to make subjective judgments. According to the similarities between the edge position of the edge image and the standard edge image, edge similarity parameters are defined to characterize the edge preserving ability of the filtering algorithm. This parameter is applied to the comparison of Gaussian noise image and salt-and-pepper noise image filtering algorithm, and the expected result is obtained. The parameter value accurately reflects the edge preserving ability of each filtering algorithm. For the lidar image, the edge preserving ability of the speckle noise filtering algorithm is of great significance. Therefore, the edge similarity is used to compare the edge preserving ability of the speckle noise filtering algorithm. The results show that the Lee filtering algorithm and the enhancement Lee The filtering algorithm has strong edge preserving ability, while the conventional average filtering and median filtering edge preserving ability is poor.