Suppr超能文献

基于正交网格条纹的条纹反射技术的快速时间相位展开方法

Fast temporal phase unwrapping method for the fringe reflection technique based on the orthogonal grid fringes.

作者信息

Li Bo, Ma Suodong, Zhai Yang

出版信息

Appl Opt. 2015 Jul 10;54(20):6282-90. doi: 10.1364/AO.54.006282.

Abstract

In traditional temporal phase unwrapping (TPU) algorithms, wrapped phases with different spatial frequencies are obtained from several groups of phase shift fringes to calculate the unwrapped phase. Therefore, the necessary quantity of captured fringes is very large, especially for the fringe reflection technique (FRT), since a pair of phases should be unwrapped to get the slopes of two perpendicular directions. In this paper, we propose a fast TPU algorithm based on the orthogonal grid fringes by which only one image is needed to extract the two integer phases for each frequency instead of two groups of phase shift fringes, and then they can be added into the wrapped phases separately to complete the unwrapping. There are ridge errors in the direct unwrapped phases, but they are significantly suppressed by our pseudo-phase-shift strategy without any extra captured fringes. The proposed method is robust and effective where the fringe amount used for unwrapping is only 1/4 of the previous similar algorithm and 1/6-1/8 of the traditional TPU methods. The detailed comparison of measurement time is also given, which demonstrate that the FRT measurement can be accelerated in most cases by our method. The algorithm is validated by the experiments, which still works well for the severely defocusing fringes or complex specimen.

摘要

在传统的时间相位展开(TPU)算法中,通过几组相移条纹获取具有不同空间频率的包裹相位,以计算展开相位。因此,所需采集的条纹数量非常大,特别是对于条纹反射技术(FRT)而言,因为需要对一对相位进行展开以获得两个垂直方向的斜率。在本文中,我们提出了一种基于正交网格条纹的快速TPU算法,通过该算法,每个频率仅需一幅图像即可提取两个整数相位,而无需两组相移条纹,然后将它们分别添加到包裹相位中以完成相位展开。直接展开的相位中存在脊误差,但通过我们的伪相移策略可显著抑制这些误差,且无需额外采集条纹。所提出的方法稳健且有效,用于相位展开的条纹数量仅为先前类似算法的1/4,传统TPU方法的1/6 - 1/8。文中还给出了测量时间的详细比较,结果表明我们的方法在大多数情况下都能加速FRT测量。该算法通过实验得到验证,对于严重散焦的条纹或复杂样本也能很好地工作。

文献AI研究员

20分钟写一篇综述,助力文献阅读效率提升50倍。

立即体验

用中文搜PubMed

大模型驱动的PubMed中文搜索引擎

马上搜索

文档翻译

学术文献翻译模型,支持多种主流文档格式。

立即体验