基于迭代法的同轴法X射线相衬成像的相位复原

Phase Retrieval of X-ray Phase Contrast Imaging Based on Iterative Algorithm

摘要: 相位复原问题是指在只有强度可以测量的情况下,从强度数据中提取出物体的相位信息。相位复原是同轴法X射线相衬成像的关键问题之一。本文采用误差递减法,通过限制物体的空间大小,结合测量到的强度数据,在物体的空域和频域都进行约束。通过反复迭代,进行带宽外推,最终实现物面和像面的相位复原。仿真实验证明了此迭代算法的收敛性,并且以塑料管为样品,在上海光源采集到同轴法X射线相衬成像的实验数据,对上述算法进行了验证。

Abstract: Phase retrieval problem is essential in the field of X-ray phase contrast imaging. Phase information is retrieved from measured intensity data. The error-reduction iterative algorithm is employed to recover phase, with spatial constrains and measured intensity data in the detector plane. Both in the numerical simulation and real imaging data test, the iterative algorithm is verified that it converges steadily.