Liao Yang, Xiong Yonghua, Yang Yunhong
State Key Laboratory of High Field Laser Physics, Shanghai Institute of Optics and Fine Mechanics, Chinese Academy of Sciences, Shanghai 201800, China.
School of Automation, China University of Geosciences, Wuhan 430074, China.
Sensors (Basel). 2021 Apr 2;21(7):2487. doi: 10.3390/s21072487.
This paper is concerned with auto-focus of microscopes for the surface structure of transparent materials under transmission illumination, where two distinct focus states appear in the focusing process and the focus position is located between the two states with the local minimum of sharpness. Please note that most existing results are derived for one focus state with the global maximum value of sharpness, they cannot provide a feasible solution to this particular problem. In this paper, an auto-focus method is developed for such a specific situation with two focus states. Firstly, a focus state recognition model, which is essentially an image classification model based on a deep convolution neural network, is established to identify the focus states of the microscopy system. Then, an endpoint search algorithm which is an evolutionary algorithm based on differential evolution is designed to obtain the positions of the two endpoints of the region where the real focus position is located, by updating the parameters according to the focus states. At last, a region search algorithm is devised to locate the focus position. The experimental results show that our method can achieve auto-focus rapidly and accurately for such a specific situation with two focus states.
本文关注透射照明下透明材料表面结构显微镜的自动聚焦，在聚焦过程中会出现两种不同的聚焦状态，且焦点位置位于这两种状态之间，清晰度存在局部最小值。请注意，大多数现有结果是针对具有清晰度全局最大值的一种聚焦状态得出的，它们无法为这个特定问题提供可行的解决方案。本文针对这种具有两种聚焦状态的特定情况开发了一种自动聚焦方法。首先，建立一个聚焦状态识别模型，它本质上是一个基于深度卷积神经网络的图像分类模型，用于识别显微镜系统的聚焦状态。然后，设计一种端点搜索算法，它是一种基于差分进化的进化算法，通过根据聚焦状态更新参数来获得真实焦点位置所在区域的两个端点的位置。最后，设计一种区域搜索算法来定位焦点位置。实验结果表明，我们的方法能够针对这种具有两种聚焦状态的特定情况快速、准确地实现自动聚焦。
