手机曲面玻璃缺陷检测方法研究  

Study on Defects Detection Method of Mobile Phone Curved Glass

作　　者：陈宁 郭钢祥[1,3] 郭斌 张跃[1,2] 王超 孔明 陈挺 王瑛辉 赵健铭[1,3] 吴键 CHEN Ning;GUO Gang-xiang;GUO Bin;ZHANG Yue;WANG Chao;KONG Ming;CHEN Ting;WANG Ying-hui;ZHAO Jian-ming;WU Jian(Zhejiang Institute of Metrology,Hangzhou,Zhejiang 310018,China;College of Metrological Technology and Engineering,China Jiliang University,Hangzhou,Zhejiang 310018,China;Key Laboratory of Digital Precision Measurement Technology of Zhejiang Province,Hangzhou,Zhejiang 310018,China)

机构地区：[1]浙江省计量科学研究院,浙江杭州310018 [2]中国计量大学计量测试工程学院,浙江杭州310018 [3]浙江省数字精密测量技术研究重点实验室,浙江杭州310018

出　　处：《计量学报》2023年第5期701-706,共6页Acta Metrologica Sinica

基　　金：国家市场监督管理总局科技计划项目(2020MK042、2020MK043);浙江省市场监督管理局重大科研项目(20210107);浙江省市场监督管理局雏鹰计划项目(CY2022337)。

摘　　要：由于手机曲面玻璃在生产过程中会产生一些外观缺陷,针对曲面玻璃上缺陷成像难、提取难等问题,提出了一种手机曲面玻璃缺陷检测方法。首先,对曲面玻璃平面部分图像进行形状匹配,并对匹配后的图像进行差分和形态学处理,提取缺陷特征;其次,对曲面玻璃曲边部分图像,使用基于连通域分析和面积阈值分割的缺陷提取算法;再次,对曲面玻璃R角部分图像,采用图像频域增强和对数变换的缺陷提取算法;最后,提取各部分缺陷后,计算得到各类缺陷的特征,通过特征进行缺陷分类,并将得到的缺陷数据与影像测量仪得到的数据进行对比实验。结果表明,该算法能准确提取手机曲面玻璃表面常见的划痕、污点、擦伤和气泡缺陷,并且缺陷尺寸精度测量能达到20μm。Due to some appearance defects in the production process of curved glass of mobile phone,a defect detection method of curved glass of mobile phone was proposed to solve the problems of difficult imaging and extraction of defects on curved glass.Firstly,shape matching was carried out on the curved glass plane images,and the matched images were processed by difference and morphology to extract the defect features.Secondly,a defect extraction algorithm based on connected domain analysis and area threshold segmentation was used for curved glass images.Thirdly,the image frequency domain enhancement and logarithmic transformation defect extraction algorithm were used for the R-angle image of curved glass.Finally,after extracting the defects of each part,the characteristics of all kinds of defects were calculated,and the defects were classified according to the characteristics,and the obtained defect data was compared with the data obtained by the image measurement instrument.The results showed that the algorithm can accurately extract the common scratches,stains,abrasions and bubbles on the curved glass surface of mobile phones,and the size accuracy of the defects can reach 20μm.

关 键 词：计量学 曲面玻璃 缺陷提取 图像匹配 连通域分析 图像增强 

分 类 号：TB96[机械工程—光学工程]