摘要
针对经典形态学方法在边缘检测时可去除图像噪声,但难以保留边缘细节的问题,提出一种能有效去除噪声且能准确检测图像边缘的方法。该方法首先利用大尺度的轮廓结构元素对图像进行滤波开、闭运算,接着用小尺度结构元素在进行经典形态学的膨胀、腐蚀运算后对图像进行梯度运算,从而得到含噪声图像的边缘信息。实验表明,该方法在准确检测图像边缘的同时,能够有效地去除图像中的噪声,且运算量相对较小。
Aiming at the problem that classical morphological method for edge detection can remove the image noise but is difficult to retain the details of the edge, a method that can effectively remove the noise and can accurately detect the details of edges is introduced. The image filtering open and close operation is performed with the large-scale contour structural elements, the gradient operation is conducted by the aid of the small-scale structural elements after the expansion and corrosion operation in the classical morphotogy, and then the edge of the image containing noise information is obtained. Experiments show that the method can effectively remove the image noise while detecting the image edges accurately, and the guantity of computation is relatively small.
出处
《现代电子技术》
2010年第8期93-96,共4页
Modern Electronics Technique
关键词
边缘检测
数学形态学
轮廓结构元素
噪声
edge detection
mathematical morphology
contour structural element
noise
