基于机器视觉石英晶片外观的缺陷检测方法  

A Method of Apparent Defect Inspection for Quartz Wafers Based on Computer Vision

作　　者：张铁[1] 邝健麟[1] 

机构地区：[1]华南理工大学机械与汽车工程学院,广州510640

出　　处：《机械科学与技术》2013年第10期1555-1560,共6页Mechanical Science and Technology for Aerospace Engineering

基　　金：国家863科技计划项目重点项目(200904AA043903-3);广东省教育部产学研结合项目(2010B090400259;2011B090400150)资助

摘　　要：针对石英晶片的外观缺陷检测技术的不足,提出一种基于机器视觉开发软件对石英晶片外观缺陷检测方法,对石英晶片生产过程中出现的外观缺陷进行了分析,提出了缺陷检测的流程,建立了基于几何特征的模板定位算法和基于Blob算法的缺陷分析算法,实现了对图像的获取、边缘检测、定位、识别和斑点分析,在C#环境下对其图像处理技术利用机器视觉软件进行了开发,搭建了石英晶片的外观缺陷检测实验平台,通过摄像机实现了石英晶片的外观图像的捕获,利用视觉软件获得石英晶片的外观质量,最后对此系统进行了测试。实验表明:该石英晶片外观缺陷检测方法具有检测速度快、精确度达到99.7%、匹配误差小于0.3等优点,可满足现场使用的要求。In order to overcome the disadvantages of the apparent defect inspection technology for quartz wafers, a new method based on the computer vision was proposed in this paper. By analyzing the appearance of the apparent defects of quartz wafers in manufacturing, the defect inspection process was built. Then, a template location algo- rithm based on the geometrical characteristics and a defect analysis algorithm based on the Blob algorithm were es- tablished in order to achieve the different functions of image processing, including acquisition, edge detection, lo- calization, recognition, blob analysis, and so forth. Finally, the system was developed by C#. Net based on the computer vision and the experimental platform was set up. The system achieved capturing the appearance image by using camera and utilized visual sense software to acquire appearance quality of quartz wafers. Furthermore, the system was tested by experiments. The results indicate that the method of apparent defect inspection for quartz wafers has many advantages, including the fast detection speed, the high accuracy which is up to 99.7%, the matching error which is less than 0. 3, and the feature which can satisfy the engineering requirements.

关 键 词：缺陷检测 机器视觉 模板 BLOB算法 

分 类 号：TP273[自动化与计算机技术—检测技术与自动化装置]