基于机器视觉的金刚线表面质量检测  

Surface quality detection of diamond wire based on machine vision

作　　者：刘明宇 佃松宜[1] LIU Ming-Yu;DIAN Song-Yi(College of Electrical and Information,Sichuan University,Chengdu 610065,China)

机构地区：[1]四川大学电气工程学院,成都610065

出　　处：《四川大学学报（自然科学版）》2020年第5期920-926,共7页Journal of Sichuan University(Natural Science Edition)

基　　金：中央高校基本科研业务费专项资金(20826041A4133)。

摘　　要：针对电镀金刚线质量品控难度大、成本高等问题,首次研制了一种基于机器视觉的金刚线离线表面质量检测系统.文中介绍了金刚线质量检测系统的结构组成,对用于评价金刚线质量情况的常用性能指标的相关算法实现进行了详细描述,并以平铺图的方式在一定程度上不加以验证的还原了金刚线的三维表面形貌.最后,根据设计制作实物并进行测试,结果表明,出刃率计算精度达93.2%,出刃高度计算精度达92.9%,满足金刚线质量检测评价的要求,在提高检测效率的同时降低检测难度.In view of the difficulty and high cost of quality control of electroplating diamond wire,an offline surface quality inspection system based on machine vision was developed for the first time.This paper introduces the structure and composition of the diamond wire quality inspection system,and describes in detail the implementation of the algorithms used to evaluate the diamond wire quality.Finally,according to the design and production of real objects and testing,the results show that the calculation accuracy of the cutting edge rate is 93.2%,the calculation accuracy of the cutting edge height is 92.9%,which meets the requirements of quality testing and evalua.

关 键 词：机器视觉 金刚线 表面质量检测 三维表面重建 

分 类 号：TP399[自动化与计算机技术—计算机应用技术]