机器视觉半导体分选系统的研制  

Development on Semiconductor Automatic Sorting System Based on Machine Vision Technology

作　　者：杜俊斌[1] 黄泽文[1] 黄云[1] 李耀棠[1] 叶世栋[1] 张世超[1] 范少武[1] 

机构地区：[1]中科院广州电子技术有限公司,广东广州510070

出　　处：《机电工程技术》2011年第8期64-65,148,共3页Mechanical & Electrical Engineering Technology

基　　金：广州市越秀区科技攻关与成果推广计划基金资助项目(2009-GX-038)

摘　　要：大批量混杂废旧半导体电子元件的自动分选对其后续环保再生处理有很重要的作用,目前多为人工目检,精度不高,效率低。设计了一种基于机器视觉OCR图像处理技术的自动分选系统,系统中废旧半导体电子元件的进料、图像获取、识别分类全流程计算机自动控制。经测试系统技术指标为:分选速率1～2个/秒,识别率70%以上,误识率小于2%。Automatic sorting with large quantities of mixed and used semiconductor components plays a significant role in follow-up environmental regeneration dispose,mainly through manual visual inspection at present,less precision and low efficiency.We designed a automatic sorting system which is based on machine vision OCR Image processing technology,the whole process of feeding image capture identification and classification is computerized According to the test,the technical specifications of system is: sorted rate is 1～2 per second,recognition rate is more than 70%,misclassification rate is less than 2%.

关 键 词：机器视觉 OCR 自动分选 环保再生 

分 类 号：TP273.5[自动化与计算机技术—检测技术与自动化装置]