Coherent Diffraction Imaging Algorithm Based On Electron Source

Coherent diffraction imaging(Coherent Diffraction Imaing,CDI)is a new imaging method that does not depend on optical elements.Its imaging resolution is no longer limited by optical elements but only depends on wavelength.At present,coherent diffraction imaging technology has been widely used in biomedicine,material science,engineering measurement,and other fields.The cold atom electron source with high brightness and coherence is not only a necessary condition for the realization of an X-ray light source but also can be directly used in scientific research.Compared with the optical imaging method,the wavelength of the electron source is shorter,and higher resolution can be achieved through diffraction imaging to observe the inner structure of the sample.The development of coherent diffraction imaging technology also provides a new idea for the application of electron sources.With the increase in application demand,how to process massive data quickly becomes the focus in the field of coherent diffraction imaging.Therefore,this paper studies the solution to this problem from the view of an algorithm.Starting from the diffraction theory of light,diffraction regions were elaborated on and deduced in detail.Based on the theory of the one-dimensional Kramers-Kronig relation(KK),we extended it to two-dimensional image processing.Then,based on the oversampling theory,the iterative phase retrieval algorithm is used to reconstruct and analyze the two-dimensional image.By comparing the error coefficient and the reconstruction results,we found that the initial value of iteration will directly affect the reconstruction result of the algorithm in mainstream iterative algorithms.Combining KK relations with oversampling smoothness(OSS)algorithm to optimize the reconstruction process,we presented a phase-retrieval algorithm named KK-OSS,improve the initial iteration value.Compared to error reduction algorithm,hybrid input-output algorithm and OSS algorithm,KK-OSS can improve the efficiency of the phase-retrieval algorithm and obtain faster reconstruction results in the keyhole CDI simulation.We modified the experimental data using poisson noise to increase the reliability of the reconstruction.Through the evaluation of the error coefficient,it is proven that under the same number of iterations,the reconstruction of KK-OSS is better than that of other algorithms with a random initial value.In the case of Fraunhofer diffraction,we also proved the superiority of KK-OSS.There are still some limitations and challenges in CDI: we need a high-quality coherent x-ray source for high-resolution CDI,and we look forward to maximizing the efficiency of the reconstruction.Therefore,the application of KK relations is worth studying.This paper focus on the important role of iterative algorithms in coherent diffraction imaging,which can make a full preparation for the subsequent application of cold atom electron source electron diffraction imaging.