基于机器学习的过焦扫描显微测量方法研究  

Through-focus scanning optical microscopy measurement based on machine learning

作　　者：李冠楠 石俊凯 陈晓梅 高超 姜行健 崔成君 朱强 霍树春 周维虎 LI Guan-nan;SHI Jun-kai;CHEN Xiao-mei;GAO Chao;JIANG Xing-jian;CUI Cheng-jun;ZHU Qiang;HUO Shu-chun;ZHOU Wei-hu(Institute of Microelectronics of the Chinese Academy of Sciences,Optoelectronic R&D Center,Beijing 100094,China;University of Chinese Academy of Sciences,Beijing 100049,China;China Banknote Printing Technology Research Institute Co.,LTD.,Beijing 100070,China)

机构地区：[1]中国科学院微电子研究所,光电技术研发中心,北京100094 [2]中国科学院大学,北京100049 [3]中钞印刷技术研究院有限公司,北京100070

出　　处：《中国光学（中英文）》2022年第4期703-711,共9页Chinese Optics

基　　金：国家重点研发计划(No.2019YFB2005603);清华大学精密测试技术及仪器国家重点实验室开放基金(No.TH20-01);国家自然科学基金(No.51905528);精密测试及仪器国家重点实验室开放基金(No.pilab2102)。

摘　　要：微电子机械系统(Micro-Electro-Mechanical System,MEMS)具有小型化、高集成度的特点,随着MEMS结构深宽比的不断增大,对MEMS结构尺寸的测量提出更高的要求。过焦扫描光学显微技术(Through-focus Scanning Optical Microscopy,TSOM)是一种高精度无损的光学测量方法,通过采集一组离焦图并沿扫描方向截取TSOM图像,利用库匹配的方法从中提取待测结构的尺寸信息。该方法对于纳米级结构测量有着极高的灵敏度,然而对于微米级特征尺寸存在建库困难且易受环境干扰的问题。本文针对微米级MEMS沟槽结构,在传统的光学显微镜基础上进行改造,建立了TSOM光学系统采集离焦图像,利用图像特征提取方法生成TSOM特征向量集,结合机器学习的方法建立不同槽宽尺寸的回归预测模型,对微米级MEMS槽宽尺寸实现纳米级测量精度,单点重复性测量2μm槽宽的相对标准差(Relative Standard Deviation,RSD)在1%左右,10μm和30μm槽宽RSD分别低于0.2%和0.35%,结果表明该方法对于微米级MEMS沟槽测量具有极高的应用前景。Micro-Electro-Mechanical Systems(MEMS)have the characteristics of miniaturization and high integration.As the high aspect ratio of MEMS increases,the measurement of MEMS feature size faces greater challenges.Through-focus Scanning Optical Microscopy(TSOM)technology is a high-precision and nondestructive optical measurement method.TSOM images are captured along the scanning direction by collecting a set of defocused images and the size information of the structure is extracted from TSOM images by the library matching method.This method is highly sensitive and suitable for nano-scale structure measurements,but it is difficult to build a database for micron-scale features and is susceptible to environmental interference.In this paper,a TSOM optical system is established and traditional optical microscopy is used to collect a set of defocused images.The TSOM’s feature vector set is obtained by the image feature extraction method and is combined with machine learning to establish MEMS groove regression prediction models with different feature sizes.The results show that the above method can achieve nano-scale high precision measurement of a MEMS groove width and the single point repeatability measurement has great performance.The Relative Standard Deviation(RSD)of 2μm width is about 1%,and the RSD of 10μm and 30μm width are respectively lower than 0.2%and 0.35%.This method has very high application prospects for micron MEMS groove structure measurement.

关 键 词：MEMS 机器学习 过焦扫描光学显微法 微纳测量 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程] P207.1[自动化与计算机技术—控制科学与工程] TH744[天文地球—测绘科学与技术]