摘要
铟锡氧化物(Indium Tin Oxide,ITO)导电层是触控显示技术的绝对定位元件,为保障定位的灵敏性和准确性,需对导电层表面缺陷进行质量检测.本文针对导电层透明区域机器视觉自动检测存在的问题,提出了图案对比度增强的方法.该方法首先利用ITO材料的光谱属性及其表面光学特性,设计出用于ITO导电层检测的近红外同轴光照明,将图像对比度从零提高到4.5%.在通过光学方法实现了对比度从无到有的转变后,充分利用数字图像预处理的优势,结合基于小波变换的非线性增强方法,最终成功将对比度提高至16%,为后续ITO导电层缺陷的分析和识别提供了良好保障.
ITO conductive layer is the absolute location components in touch display technology. It is necessary to inspect the surface defects of ITO layer for guaranteeing the sensitivity and accuracy of touch operation. According to the issue existing in inspecting transparent area of ITO layer automatically based on machine vision technology, a method used for enhancing contrast of pattern on ITO layer is presented in paper. The method exploited the spectral property of ITO and its interaction with light, and a NIR coaxial light was designed to illuminate. The optical method improved the contrast of ITO pattern from 0 to 4.5%. Based on the fundamental contribution of illumination, nonlinear enhancement performed by wavelet transform was further applied. The contrast eventually increased to 16 %, which provides a good reliability to analyze and recognize ITO layer defects.
出处
《光子学报》
EI
CAS
CSCD
北大核心
2015年第5期194-198,共5页
Acta Photonica Sinica
基金
广东省引进创新科研团队计划(No.201001D0104799318)资助
关键词
机器视觉
小波变换
LED照明
表面缺陷
图像增强
透明导电薄膜
缺陷识别
Machine vision
Wavelet transform
LED illumination
Surface defects
Image enhancement Transparent conductive film
Defect recognition