摘要
以智能制造业表面缺陷在线自动检测为应用背景,系统地综述了自动光学(视觉)检测(以下统称自动光学检测,AOI)技术。内容涉及AOI技术的基本原理、光学成像方法、系统集成关键技术、图像处理与缺陷分类方法等。对AOI系统集成中的关键技术,如视觉照明技术、大视场高速成像技术、分布式高速图像处理技术、精密传输和定位技术和网络化控制技术等进行了概述;对表面缺陷AOI主要光学成像方法的基本光学原理、功能和应用场合进行了总结;对表面缺陷检测中的图像处理、缺陷几何特征定义、特征识别与分类算法进行了系统阐述,重点介绍了周期纹理表面缺陷图像中的纹理背景去除方法,复杂和随机纹理表面缺陷的深度学习检测、识别与分类方法。
The authors comprehensively review technique of automated optical(visual)inspection(AOI)technique from aspects of the basic principle,optical imaging method,key techniques of system integration,image processing and defect classification at the application background of automated online surface defect inspection in intelligent manufacturing industry.The key technologies of system integration in automated optical inspection,such as visual lighting,high speed imaging in a large field of view,distributed high-speed image processing,precision transmission and positioning for the inspected objects,and networked control,are briefly summarized.The basic optical principles,functions and applications of the optical imaging methods commonly used in automated optical defect inspection are comprehensively reviewed.The image processing,defect geometric feature definition,feature recognition and classification algorithm for surface defect inspection are systematically summarized.Particularly,the methods of texture background removal in the images with periodic textures,and the detect detection,recognition and classification methods for complex and random texture surface based on depth learning are reviewed.
作者
卢荣胜
吴昂
张腾达
王永红
Lu Rongsheng;Wu Ang;Zhang Tengda;Wang Yonghong(Schoot of Instrument Science and Opto-Etectronics Engineering,Hefei University of Technology,Hefei,Anhui 230009,China;College of Mechanical and Electrical Engineering,Henan Agricultural University,Zhengzhou,Henan 450002,China)
出处
《光学学报》
EI
CAS
CSCD
北大核心
2018年第8期15-50,共36页
Acta Optica Sinica
基金
国家重大科学仪器开发与应用专项(2013YQ220749)
国家重点研发计划(2016YFF0101803)
关键词
机器视觉
表面缺陷
自动光学检测
视觉检测
图像处理
分类
machine vision
sur{ace defect
automated optical inspection
vision inspection
image processing
classification