基于光谱估计与多光谱技术的光学元件表面疵病检测  

Measurement of Surface Defects of Optical Elements Based on Spectral Estimation and Multispectral Technique

作　　者：杨言若 步扬[1] 徐静浩[1] 王少卿[1] 王向朝[1] 李杰 Yang Yanruo;Bu Yang;Xu Jinghao;Wang Shaoqing;Wang Xiangzhao;Li Jie(Laboratory of Information Optics and Optical Technology , Shanghai Institute of Optics and Fine Mechanics ,Chinese Academy of Sciences , Shanghai 201800, China;Center of Materials Science and Optoelectronics Engineering , University of Chinese Academy of Sciences ,Beijing 100049, China;Research Center of Laser Fusion , China Academy of Engineering Physics , Mianyang , Sichuan 621000, China)

机构地区：[1]中国科学院上海光学精密机械研究所信息光学与光电技术实验室,上海201800 [2]中国科学院大学材料与光电研究中心,北京100049 [3]中国工程物理研究院激光聚变研究中心,四川绵阳621000

出　　处：《中国激光》2019年第9期153-161,共9页Chinese Journal of Lasers

基　　金：国家自然科学基金(61405210);国家科技重大专项(2009ZX02205-001,2016ZX02201-001);广西高校光电信息处理重点实验室开放基金(KFJJ2016-03)

摘　　要：为实现精密光学元件表面疵病的高效测量和精确统计,提出了一种基于光谱估计和多光谱技术的光学元件表面疵病检测方法。该方法利用光谱估计提取白光图像中不同波长的单光谱疵病图像,并合成多光谱疵病图像,然后采用优化后的OTSU(Otsu Image Segmentation Algorithm)分别对单光谱与多光谱疵病图像进行分析。基于该方法搭建了光学元件表面疵病检测装置,获得了白光照明条件下光学元件表面疵病的图像。实验结果表明,与原始白光图像相比,合成多光谱图像的疵病检出数量提升了1.85倍,疵病检出面积最大增加了6.0倍,检测效率得到明显提高。根据光学元件表面疵病的特性选取不同波长组合来生成单光谱与多光谱图像,可更加高效精确地检测出传统检测技术不易检出的疵病信息。A method based on spectral estimation and a multispectral technique is proposed to accurately and efficiently measure surface defects of optical elements. Using this method, we estimate and obtain single spectral defect images at different wavelengths and compose a multispectral defect image from different single spectral defect images. The optimized OTSU (Otsu Image Segmentation Algorithm) is used to analyze and compare the images. A defect measurement system of optical elements is developed, and color defect images are obtained upon white-light illumination. The experimental and analytical results show that single spectral or composing multispectral images from different wavelength combinations detect more defects accurately and efficiently. In the composite image, the number of detected defects increases by 1.85 times, and the area of detected defects increases by approximately 6.0 times for different kinds of defects compared to the original white-light image. The composing images from optimized wavelength combi nations can be used to detect and obtain different small defects with high efficiency and high quality.

关 键 词：测量 疵病检测 光谱估计 多光谱技术 疵病数量 疵病面积 

分 类 号：O439[机械工程—光学工程]