液晶与显示2016，Vol.31Issue(6)：613-620,8.DOI:10.3788/YJYXS20163106.0613
基于改进Chan-Vese模型的液晶显示屏Mura缺陷分割
LCD Mura defectimage segmentation based on improved Chan-Vese model
摘要
Abstract
LCD Mura defect is difficult to detect.It has the features of low contrast,uneven bright-ness,blurry edge,and so on.The traditional Chan-Vese model (C-V model)cannot accurately seg-ment the Mura defect and the segmentation speed is slow.In order to solve the problems,this paper presents an improved C-V model.First of all,according to the theory of curve evolution,the image data force in the traditional C-V model is simplified to reduce the iterative process of calculation and improve the segmentation speed.Secondly,in order to reduce the effect from non-uniformity bright-ness of image,the model is proved by adding a new energy term that related to the brightness differ-ence between inside and outside area of the contour curve to improve the accuracy of the segmentation. Finally,the iterative stopping criterion is introduced in the implementation process of the algorithm, the setting of the segmentation accuracy can aid both in achieving the automatic iteration stop and seg-menting the target correctly.Experimental results indicate that improved C-V model in this paper can accurately segment the background uneven Mura defect and it has a faster speed.关键词
Chan-Vese模型/液晶显示屏 Mura缺陷/亮度不均匀Key words
Chan-Vese model/LCD Mura defect/uneven brightness分类
信息技术与安全科学引用本文复制引用
陈凌海,姚剑敏,郭太良..基于改进Chan-Vese模型的液晶显示屏Mura缺陷分割[J].液晶与显示,2016,31(6):613-620,8.基金项目
国家“863”重大专项(No．2013AA030601) （No．2013AA030601）
