Since the classic edge detection algorithm is involved in the computation of the gradient, is sensitive to noise, and has a large amount of computation, the SUSAN algorithm that is only based on compared neighboring pixel grayscale, but is not involved in gradient operation at all is proposed to realize an effective edge detection which has a small amount of calculation. The SUSAN algorithm is compared with other main edge detecting algorithms such as the Prewitt algorithm, the Gauss-Laplace algorithm and the Canny algorithm. The Matlab simulation experiments indicate that the SUSAN algorithm does not need to calculate the differential, has a small amount of calculation and reduces the noise impact. The selection of the SUSAN algorithm to control parameters is simple, so it is easy to achieve automation selection. It is a very effective algorithm.%针对经典的边缘检测算法均涉及梯度的运算,存在对噪声敏感、计算量大等缺点的问题,为了找到一种检测效果好、计算量相当的图像边缘算法,提出了只基于对周边像素的灰度比较,完全不涉及梯度运算的SUSAN算法.主要研究了SUSAN算法,并比较了主流的Prewitt算子,Gauss-liaplace算子,Canny算子的边缘检测算法,经Matlab仿真实验后,确定该算法不需要计算微分,计算量小,进一步减少了噪声的影响,且SUSAN算子控制参数的选择很简单,任意性较小,所以比较容易实现自动化的选取,是一种非常有效的算法.
展开▼
