相(物质)
小波
小波变换
算法
奇点
计算机科学
非线性系统
相位恢复
噪音(视频)
迭代法
相位展开
人工智能
图像(数学)
计算机视觉
光学
傅里叶变换
数学
数学分析
物理
干涉测量
量子力学
作者
Chen Fan,Hong Zhao,Zixin Zhao,Junxiang Li,Yijun Du,Gaopeng Zhang
出处
期刊:IEEE Transactions on Instrumentation and Measurement
[Institute of Electrical and Electronics Engineers]
日期:2023-01-01
卷期号:72: 1-11
被引量:3
标识
DOI:10.1109/tim.2023.3280499
摘要
We propose a high accurate quantitative phase imaging (QPI) method by using transport of intensity equation (TIE) and wavelet transform (WT). TIE provides a simple and fast method for QPI, but its accuracy is always limited due to the nonlinear error and noise problems caused by the defocus distance. To improve the accuracy of the phase recovered by TIE, WT is introduced to combine with TIE under several defocus distances. With the help of the multiresolution characteristics of the WT, effective information can be extracted from the phases retrieved with TIE at different defocus distances. As a result, a more accurate phase can be obtained by fusing this effective information. Moreover, to extend the applicability of our method, the problems of phase discrepancy and phase singularity in TIE are discussed and solved with an iterative WT-TIE algorithm. Numerical simulations and experiments with various types of phase maps are presented to comprehensively demonstrate the accuracy and effectiveness of our method.
科研通智能强力驱动
Strongly Powered by AbleSci AI