基于数学形态学和LMS滤波的图像边缘检测算法研究

摘要

传统的边缘检测算子对灰度图像进行边缘检测时存在图像细节被丢失,边界不连续等问题.针对上述问题,提出一种基于数学形态学和最小均方差滤波相结合的图像边缘检测方法,该算法先利用小均方差滤波的方法可以有效地滤除图像中的噪声,然后利用形态学中的腐蚀运算对图像进行边缘检测处理.实验结果表明:该方法能够有效地去噪,精确地检测图像中的细节,并且边界的连续性好.%For gray image, the traditional edge detection method has the problem of losing image detail and has discontinuous image boundary. In order to solve these problems, a edge detection algorithm based on mathematical morphology and least mean square filter is proposed. First LMS filtering is used to decrease the noise of image, then using corrosion process of morphological to detect the edge of image. The experimental results show that this edge detection method can effectively get rid of noise and correctly detect the continuous image edge.