基于机器视觉的玻璃缺陷检测系统的研究-光学工程专业论文.docxVIP

ABSTRACT DMD， as a critical component in projector， is packed inside an enclosed space built by metal and glass. Optical glass cover has characteristics of high light transmittance and high hardness. Machine vision inspection ， as a brand-new measurement technique， with characteristics of good speed， high precision， noncontact and stable performance， brings revolutionary change to glass production enterprises. By analyzing components of existing detection system， lighting unit， image collection and processing unit， combine actual demand， finish modification on stage， detection fixture， CCD and image colleting card of original detection system. At the same time， on the basis of research about glass defect image processing and grey image detection principle， use VC6.0 to design image processing program. Finally， by system selecting， assembling and testing， complete improved glass defect detection system for T-A products. New detection system has key functions of glass image collecting and processing， image defect analysis. Through computer control， whole detection system adopts high speed CCD camera to capture glass images， through image grabbing card in PVC system to “understand” image and transfers to system recognizable language and processes image by using SPC system， then gets final output on glass defects. Mass experiment data can prove that this design is feasible. The detection speed and precision meets desired design requirements， meet inspection demand at work site and guarantee prompt shipment of qualified products KEY WORDS：Machine Vision Glass Defect CCD 目 录 第一章 绪论1 1.1 本文背景及研究意义 1 1.2 国内外研究现状 2 1.3 论文研究的主要内容 3 1.4 本章小结 3 第二章 总体方案设计4 2.1 玻璃缺陷检测技术分析 4 2.2 缺陷检测系统的设计 4 2.2.1 光照单元设计6 2.2.2 图像采集单元设计10 2.2.3 图像处理单元设计16 2.3 本章小结 17 第三章 玻璃缺陷图像处理18 3.1 图像形式的介绍 18 3.2 图像格式的介绍 19 3.3 图像消噪 21 3.4 图像分割 24 3.5 本章小结 29 第四章 图像灰度的检测研究30 4.1 图像特征提取原理 30 4.2 玻璃缺陷的特征 31 4.3 玻璃缺陷的特征提取 31 4.4 缺陷点判断 33 4.5 产品符合性判定 33 4.6 本章小结 33 第五章 实验数据分析34 5.1 数据分析