摘要
对TFT-LCD表面缺陷的检测方法做了综述。首先介绍了TFT-LCD表面缺陷的种类,分析了缺陷的特点和成因,然后介绍了缺陷检测方法:图像识别法和图像处理法。在图像识别法中,详细分析了缺陷图像降维、缺陷特征提取和分类器设计等关键技术。在图像处理法中,探讨了边界模糊缺陷分割法、差影法和滤波法。最后归纳总结了各种检测方法的优点和不足,并指出了TFT-LCD表面缺陷检测的未来研究方向。
Thin Film Transistor Liquid Crystal Display （TFT-LCD） surface defect detection methods are summarized. First, types of the defect are introduced, and characters and causes are analyzed. Then two types of defect detection methods, image recogni- lion and image processing, are inlroduced. Some key technologies of reducing dimensions, feature extraction and classifier design are discussed in detail in image recognition. Some methods of segmenting defects of vogue boundary, image subtraction and filte- ring are analyzed in image processing. Finally advantages and disadvantages of various detection methods are concluded, and some issues for the next research are also pointed out.
出处
《电视技术》
北大核心
2015年第9期146-152,共7页
Video Engineering
基金
国家基础研究"973"项目(2011CB013104)
国家自然科学联合基金重点合作项目(U1134004
51275093)
高等教育学科建设重点项目(2012CXZD0020)
广东省专项研发项目LED行业基金项目(2011A081301001
2012A08030300)
省部产学研结合项目(2012B091000028)
数控一代项目(2012B011300027)
关键词
缺陷检测
特征提取
分类器
差影
detect defection
feature exlraetion
classifier
image subtraction
