首页|TFT-LCD缺陷检测系统的研究

TFT-LCD缺陷检测系统的研究

扫码查看
随着TFT-LCD的技术飞速发展及广泛应用,人工检测良品率的方法已不能保证检测效率和检测精度.针对此问题,提出了一种基于机器视觉理论的TFT-LCD缺陷检测系统,用以满足工业自动化的需求.该系统搭建面阵CCD采集图像平台,检测算法包括提取图像的ROI区域、Gabor滤波、自适应二值化、连通域提取缺陷、计算缺陷的属性特征并标记轮廓线等步骤,最后得出结果.实验结果表明,检测系统对点缺陷、线缺陷的识别率非常高,Mura缺陷针对正视可见的效率良好.提出了一种耦合性好的检测方法,针对不同种类的缺陷采用同一种检测技术,简明、方便、准确地实现了缺陷检测,并可实际应用到工业自动化中.
Research on detection system for TFT-LCD defects
With the rapid development of TFT-LCD technology,manual detection methods cannot guarantee the detection efficiency and detection accuracy.Aiming at this problem,a TFT-LCD defect detection system based on ma chine vision is presented in order to satisfy the demand of industrial automation.The system establishes the area array CCD image acquisition platform.The detection algorithm includes extraction of ROI image,Gabor filtering,adaptive two valued attributes,connected region extraction of defect,feature calculation of defect and mark the outline steps,the final outcome.The experimental results show that the recognition rate of the detection system for point defect and the line defect is very high,and the recognition rate for the visible Mura defect in face is good.A coupling technique for defect detection proposed in this paper,according to the different kinds of the same kind of detection technology,realizes simple,convenient and accurate defect detection,and can be applied to the industrial automation.

TFT-LCDalgorithmGabor filteringMura

王新新、徐江伟、邹伟金、刘永丰、王秀丽

展开 >

北京兆维电子(集团)有限责任公司 北京100015

河北工业大学机器人及自动化研究所 天津300130

TFT-LCD 算法 Gabor滤波 Mura

北京市科学技术委员会基金

Z121100007812001

2014

电子测量与仪器学报
中国电子学会

电子测量与仪器学报

CSTPCDCSCD北大核心
影响因子:2.52
ISSN:1000-7105
年,卷(期):2014.28(3)
  • 32
  • 7