一种基于改进算子的形态学边缘检测算法

An Improved Edge Detection Algorithm Based on Morphologic Operators

针对传统的边缘检测算法存在的不足,文中基于修正的边缘检测算子和形态滤波思想,采用多尺度和多结构元素,提出了一种新的多结构多尺度形态学灰度图像边缘检测算法,通过构造新算子实现边缘检测。在该算法中,分别将各个结构元素下的检测结果进行加权求和,得到图像的边缘。实验表明,和其他的传统或形态学边缘检测算法相比,文中方法具有更好的噪声抑制能力,而且边缘定位准确,检测到的边缘轮廓更加清晰完整。

For the problem of the traditional edge detection algorithm, on the basis of the improved edge detection operator and the idea of morphological filter ,propose a new multi-scale morphological edge detection algorithm of gray image, through the construction of the new operator realize edge detection. The ideal image edge is obtained by integrating the detection result of each element. The experimental results show that compared with other traditional or morphological edge detection methods, the proposed method has better noise suppression ability, and edge location is accurate, detected edge profile is more clear and complete.