改进Canny算法的热管焊头缺口边缘检测研究  

Research on Notch Edge Detection of Heat Pipe Welding Head Based on Improved Canny Algorithm

作　　者：沈光辉 杨龙兴 Shen Guanghui;Yang Longxing(School of Mechanical Engineering,Jiangsu University of Technology,Changzhou,Jiangsu 213001,China)

机构地区：[1]江苏理工学院机械工程学院,江苏常州213001

出　　处：《机电工程技术》2022年第4期162-166,共5页Mechanical & Electrical Engineering Technology

摘　　要：为了解决传统热管折弯作业模式存在的弊端,提高热管折弯精度,实现热管折弯工序的自动化,设计了一种基于机器视觉的热管检测加工一体化设备。首先介绍该设备的系统结构与电气控制,并分析其运动精度,然后为了提高热管焊头缺口边缘图像的采集质量,从而保证定位精度与加工精度,在传统的Canny算法基础上进行了两方面的改进,一是利用双边滤波代替传统高斯滤波平滑图像,在滤波消噪的同时增强对边缘轮廓的保护;二是采用遗传算法优化OTSU法对非极大值抑制后的梯度图像进行分割阈值求取,实现Canny边缘提取的自适应。实验结果表明,该算法能够实现Canny双阈值选取过程的自适应,且可在去除非目标信息的基础上较完整地检测出边缘轮廓,满足热管检测加工的精度要求。In order to solve the disadvantages of the traditional heat pipe bending operation mode,improve the heat pipe bending accuracy and realize the automation of heat pipe bending process,an integrated heat pipe detection and processing equipment based on machine vision was designed.Firstly,the system structure and electrical control of the equipment were introduced,and its motion accuracy was analyzed.Then,in order to improve the acquisition quality of the notch edge image of the heat pipe welding head,so as to ensure the positioning accuracy and machining accuracy,two improvements were made based on the traditional Canny algorithm.One was to smooth the image by using bilateral filter instead of traditional Gaussian filter,the edge contour protection was enhanced while filtering and denoising.Secondly,genetic algorithm was used to optimize OTSU method to calculate the segmentation threshold of gradient image after non maximum suppression,so as to realize the adaptive Canny edge extraction.The experimental results show that the algorithm can realize the adaptation of Canny double threshold selection process,and can completely detect the edge contour on the basis of removing non target information,which can meet the accuracy requirements of heat pipe detection and processing.

关 键 词：热管焊头缺口检测 改进Canny算法 遗传算法 自适应 

分 类 号：TP399[自动化与计算机技术—计算机应用技术]