一种适应多方向灰度梯度变化的自动对焦评价函数  

An Auto-Focusing Evaluation Function Adapted to Multi-Directional Gray Gradient Change

作　　者：熊锐 顾乃庭[2,3] 徐洪艳 Xiong Rui;Gu Naiting;Xu Hongyan(Quality Inspection Center,Institute of Optics and Electronics,Chinese Academy of Sciences,Chengdu,Sichuan 610209,China;Key Laboratory of Adaptive Optics,Institute of Optics and Electronics,Chinese Academy of Sciences,Chengdu,Sichuan 610209,China;University of Chinese Academy of Sciences,Beijing 100049,China)

机构地区：[1]中国科学院光电技术研究所质检中心,四川成都610209 [2]中国科学院光电技术研究所自适应光学重点实验室,四川成都610209 [3]中国科学院大学,北京100049

出　　处：《激光与光电子学进展》2022年第4期365-372,共8页Laser & Optoelectronics Progress

基　　金：国家自然科学基金(12022308)。

摘　　要：当利用传统聚焦评价函数对微纳结构进行自动对焦时,微纳结构边缘方向的多样性会导致对焦稳定性差和灵敏度低。针对上述问题,对比分析了传统的聚焦评价函数,并将Brenner函数和Roberts函数相结合,提出了一种新的聚焦评价函数——Brenner2d_Roberts函数。通过实验获取了三组不同边缘方向的微结构显微图像,并对生成的对焦评价曲线及相关的评价指标进行了对比分析。结果表明,与SMD、Brenner和Roberts等传统聚焦评价函数相比,所提聚焦评价函数在多个显微图像边缘方向上均表现出良好的对焦性能,且具有更好的稳定性和更高的灵敏度。When the traditional focus evaluation function is used to automatically focus the micro-nano structure,the diversity of the edge direction of the micro-nano structure will lead to poor focus stability and low sensitivity. In view of the above problem, the traditional focus evaluation functions are compared and analyzed, and a new focus evaluation function, i. e., Brenner2 d_Roberts function is proposed by combining the Brenner function and Roberts function. Three groups of microstructure images with different edge directions are obtained through experiments,and the focus evaluation curves and related evaluation indexes are compared and analyzed. The results show that,compared with the traditional focus evaluation functions such as SMD, Brenner and Roberts, the proposed focus evaluation function shows better focus performance in multiple edge directions of microscopic images, and has better stability and higher sensitivity.

关 键 词：显微 自动对焦 显微图像 聚焦评价函数 稳定性 

分 类 号：O439[机械工程—光学工程]