针对椒盐噪声图像的Canny边缘检测算法

Canny edge detection algorithm for images with salt and pepper noise

针对传统Canny算子对椒盐噪声敏感的问题,提出一种针对椒盐噪声的Canny边缘检测算法。采用反距离加权插值滤波器替换传统Canny算子中的高斯滤波器;并在非极大值抑制部分采用非噪声点优选原则。仿真实验证明,本文算法在边缘检测时,能有效地避免椒盐噪声的干扰,即使椒盐噪声强度达到0.8,仍然可以保持漏检率和误检率在很低的程度,是一种有效的针对椒盐噪声的边缘检测算法。

In view of the problem that the traditional Canny operator is sensitive to salt and pepper noise,a Canny edge detection algorithm for salt and pepper noise is proposed.The inverse distance weighted interpolation filter is used to replace the Gauss filter in the traditional Canny operator,and the principle of non-noise spot selection is adopted in the non-maximum suppression part.Simulation experiments show that the proposed algorithm can effectively avoid the interference of salt and pepper noise in edge detection.Even when the intensity of salt and pepper noise reaches 0.8,the missed detection rate and the false detection rate can still be kept to a low level,which is an effective edge detection algorithm for salt and pepper noise.