基于变焦技术的三维微纳表面形貌测量算法研究

2018-07-16

作者：尹秋霞1，史艳琼2，卢荣胜1 单位：1合肥工业大学；2淮南联合大学

摘要：利用变焦原理，通过控制步进电机的运动，不断改变成像CCD与被测物体的z向间距，拍摄获取一维、二维图像序列。通过比较分析各种聚焦评价和搜索算法，采用抗噪性和处理速度较好的Tenengrad聚焦评价函数和高斯曲线拟合搜索方法，通过空间域的像素融合得到被测物体的全景深图像。实验结果表明，采用上述算法能够较好地重现被测物体的三维结构，纵向分辨率能够满足测量精度要求。

关键词：数字图像相关；聚焦评价函数；极值搜索算法；图像融合

中图分类号：TG806；TH124文献标志码：AAlgorithm Study on 3D Micronano Surface Topography Measurement Based on Variation Focus TechnologyYin Qiuxia，Shi Yanqiong，Lu Rongsheng

Abstract:In this paper，by using the variation focus principle and controling the movement of stepper motor to change the distance between imaging CCD and the object continuously，a sequence of twodimensional images are obtained.By comparing various focusing evaluation functions and searching algorithm，adopting tenengrad focusing evaluation function and Gauss curve fitting search method is decided，which act better in noise immunity and processing speed.Through the fusion of pixels in the spatial domain，the panoramic image of the object is obtained.The experimental result shows that the proposed algorithm can reproduce the 3D structure of the object，and the vertical resolution can meet the measurement accuracy requirements.

Keywords:DIC；focusing evaluation functions；extremum searching algorithm；image fusion