基于改进Canny算法的工件边缘检测方法

Workpiece edge detection method based on improved Canny algorithm

为了使区域模板匹配定位能够更准确地进行,需要机器视觉精准地检测出工件边缘信息,同时消除噪声、伪边缘等无关信息的干扰。针对采用Canny算法进行边缘检测时图像平滑处理过度、阈值需要预先确定、对无关信息消除能力较弱等缺点,提出一种基于改进Canny算法的工件边缘检测方法。采用双边滤波方法处理图像,以有效去除噪声并保留边缘信息;增加45°与135°方向的梯度模板计算梯度幅值,使更多的边缘信息能够检测出来;运用最大类间方差法来确定Canny算法的高阈值,使边缘检测算法具有较强的自适应能力。实验结果表明,改进的算法能够准确地检测到真实边缘,具备较强的自适应性,特别是对存在划痕的工件进行边缘检测时,此方法能够消除划痕干扰,其检测效果优于传统Canny算法。

In order to make the regional template matching and positioning more accurately,for machine vision,to precisely detect the workpiece edge and remove the interference of irrelevant information,such as noise and spurious edges are need. To solve the shortcomings such as excessive image smoothing,thresholds need to be preliminarily determined,and weak elimination of irrelevant information by Canny algorithm,a method of edge detection based on improved Canny algorithm is proposed. In order to effectively remove the noise and preserve the edge information,the bilateral filtering method is used for image processing while gradient templates are increased by 45 degrees and 135 degrees direction calculation of the gradient amplitude. So the more edge information can be detected; and the maximum interclass variance method is used to determine the high threshold ofCanny algorithm,so that the edge detection algorithm has better adaptive ability. The experimental results show that the improved algorithm can detect the true edge accurately,and has strong adaptability. Especially when the edge detection of the workpiece is scratched,this method can eliminate the scratches interference,the detection effect is better than the traditional Canny algorithm.