边缘检测在计算机视觉几何测量中的应用

在介绍边缘检测的主要特征的基础上，分析了目前几种典型图像边缘检测算法。针对计算机视觉在几何测量应用中边缘检测特点，提出了将高斯滤波与一阶导数相结合的一种新的边缘检测方法，并开发了相应的软件。其基本方法是：首先用高斯滤波器平滑图像；再用一阶偏导的有限微分算子来计算梯度的方向和幅值；对梯度幅值进行非极大值抑制；用双阈值算法检测和连接边缘。实验验证该方法提高了对被测工件边缘敏感性的同时较好地抑制了噪声。

Edge Detection Based on computer vision in geometrical measurement

Base on the characteristics of edge detection,several typical edge detection methods are analyzed in detail.Methods for edge detection on computer vision in geometrical measurement and the software design of the system are stated.The method can be summarized as follows: First the image is smoothed using a Gaussian filter,then the local gradient and edge direction are computed at each point,then the edge points determined give rise to ridges in the gradient magnitude image,finally,the algorithm performs edge linking.The experiment results : In addition to detecting the desired features,the method also produced the cleaner map.