Image reconstruction method and system based on laminated self-focusing experiment

Abstract

The invention discloses an image reconstruction method and system based on a laminated self-focusing experiment. The method comprises the following steps: placing a to-be-detected sample on an x-y two-dimensional mechanical translation table, and collecting diffraction data of all translation positions of the to-be-detected sample; based on the diffraction data, performing initial laminated imaging numerical reconstruction to obtain a reconstructed image; calculating image quality evaluation operators of all reconstructed images within a preset distance range around the initial diffraction distance; searching a maximum value of the image quality evaluation operator by using an Adam optimization method, and determining a gradient updating direction; and updating the diffraction distance and the reconstructed image according to the gradient updating direction until the number of iterations is reached. A TV image quality evaluation operator based on a self-adaptive Lp norm solves the problem of inaccurate distance result convergence caused by a complex texture structure of a biological sample in a conventional self-focusing algorithm, and an optimal diffraction distance is obtained, so that the image quality of a reconstructed image is improved.

Description

technical field [0001] The present invention relates to the technical field of lamination imaging, in particular to an image reconstruction method and system based on lamination self-focusing experiments. Background technique [0002] Ptychography is a mainstream phase imaging method. The general operation steps of traditional ptychography are: 1. Use a beam of probe beam to irradiate the sample; The illumination area has a certain overlap rate; 3. Record the diffraction data of all moving positions at a certain diffraction distance from the sample; 4. Perform digital reconstruction based on the collected diffraction data and some prior information. In this process, the diffraction distance z between the sample and the detector (CCD) is very important prior information, and the accuracy of this parameter will greatly affect the quality of the reconstructed image. Manually measuring the z value when using plane wave irradiation is often not accurate enough, and often needs t...

AI Technical Summary

Problems solved by technology

In this process, the diffraction distance z between the sample and the detector (CCD) is very important prior information, and the accuracy of this parameter will greatly affect the quality of the reconstructed image

Images