基于机器视觉的微型电子接插件尺寸测量
更新日期:2021-06-15     浏览次数:142
核心提示:摘要为了解决传统尺寸测量方法在微型电子接插件测量上准确度偏低的问题,对传统尺寸测量方法进行了改进。在分析传统测量方法的流程、算法适用性的基础

摘要 为了解决传统尺寸测量方法在微型电子接插件测量上准确度偏低的问题,对传统尺寸测量方法进行了改进。在分析传统测量方法的流程、算法适用性的基础上,通过增加规则边缘拟合,调整模板匹配的顺序,对测量流程上进行了改进;在边缘检测过程中,对传统Canny算法进行改进,提出了一种自适应阈值非闭合弱边缘剔除的边缘检测算法。所提出的测量方法有效降低了外部环境造成的测量偏差。通过搭建实验软硬件平台,并以BK-3.0微型电子接插件为被测对象进行验证。测量结果表明,所提出的方法比传统检测方法在准确度上提高了32%,从而验证了该方法的测量准确性。 In order to solve the problems of slow positioning and low stability of traditional dimension measurement method in measuring micro-electronic connector,the traditional size measurement method is improved.based on the analysis of the process and applicability of traditional measurement methods,the measurement process is improved by adding regular edge fitting and adjusting the order of template matching.In the process of edge detection,the traditional Canny algorithm is improved,and an edge detection algorithm of adaptive threshold non-closed weak edge elimination is proposed.The proposed measurement method can effectively reduce the measurement deviation caused by external environment.By building an experimental software and hardware platform,and taking BK-3.0 micro-electronic connector as the test object,the measurement results show that the accuracy of the proposed method is 32% higher than that of the traditional detection method,which verifies the measurement accuracy of the method.
作者 俞洋 金彬 YU Yang;JIN Bin(School of Electric Information Engineering,Jiangsu University of Technology,Changzhou 213001,Jiangsu,China)
出处 《实验室研究与探索》 CAS 北大核心 2021年第4期13-18,共6页 Research and Exploration In Laboratory
基金 国家自然科学基金资助项目(61601208) 江苏省自然科学基金资助项目(BK20160294)。
关键词 机器视觉 模板匹配 边缘检测 尺寸测量 machine vision template matching edge fitting dimension measurement