抗噪形态学图像边缘检测算法的研究

Research on image edge detection algorithm based on mathematical morphology

在图像的边缘检测中,影响边缘检测效果的因素有很多,其中噪声是主要因素。目前,已有的边缘检测算法在实际的检测中存在各种问题,如抗噪性差、算法复杂、运算量大、出现伪边缘等等。为了解决上述问题,提高检测的效率且较好地保留图像的边缘信息,文中提出了一种结合数学形态学的抗噪边缘检测新算法,采用两个互不相同的结构元素对图像边缘进行检测。通过实验可以验证,新算法具有抗噪能力较强、边缘定位准确且很好地保留图像边缘细节等特点。

There are many factors that influence the effect of edge detection in the image edge detection,in which the noise is the main factors.At present,the existing edge detection algorithms in the actual detection exist various problems,such as poor anti-noise,complex algorithms,large amount of computation,false edges and so on.In order to solve the above problems,to improve the detection efficiency and better retain edge information of the image,this paper proposes a combination of mathematical morphology of anti-noise edge detection algorithm,using two mutually different structure elements to detect the image edge.The experimental results show that the new algorithm has the characteristics of strong anti-noise ability,accurate edge location and good image edge preserving.