一种针对不完善数据的基于全变分约束的相干衍射算法

A modified coherent diffraction algorithm based on the total variation algorithm for insufficient data

X射线相干衍射成像(coherent diffraction imaging,CDI)技术是一种无透镜成像技术.其基本原理是使用高度相干的X射线光束照射孤立样品,在远场收集相干衍射图样的信息,使用CDI重构算法从衍射图样中还原出样品的真实结构信息.但是由于实验技术的限制,通常实验数据都是不完善的,所以相干衍射算法对于噪声和数据缺失的容忍能力是非常重要的.通过将全变分约束引入到CDI重构算法中得到了一种改进后的算法,以提升算法对噪声和数据缺失的容忍能力.并使用模拟数据与真实实验数据来进行验证,结果表明改进后算法可以加快算法的收敛速度以及提高对噪声和数据缺失的容忍能力.

X-ray coherent diffraction imaging(CDI)is a lensless imaging technology.Its basic principle is to illuminate an isolated sample with a highly coherent X-ray beam,then collect the information of the coherent diffraction pattern in the far field,and last restore the real structure information of the sample from the diffraction pattern by using the the CDI algorithm.Due to the limitation of experimental technology experimental data are usually defective,thus tolerance to noise and missing data is an important indicator for the CDI algorithm.Here a modified coherent diffraction algorithm by adding the total variation(TV)constraint into the CDI reconstruction algorithm was developed to improve the tolerance to noise and missing data.Then the performance of the modified coherent diffraction algorithm based on the total variation algorithm was verified using simulation data and experimental data.The results show the modified algorithm can accelerate convergence and improve the tolerance to noise and missing data.