Variable-angle illumination tomography method and device based on deep learning generation network

Abstract

The invention discloses a variable-angle illumination tomography method and a device based on a deep learning generation network. The method comprises the following steps of according to the Helmholtzequation of the wave and the Fourier propagation model of the light, deriving a distribution model of a diffraction field and a refraction field when the light propagates layer by layer in a non-uniform transparent medium; generating a neural network by simulating the physical process of the deep learning during the time-domain and frequency-domain propagation process in a complex form; through an angle spectrum propagation formula, backwardly transmitting a collected output optical complex field wherein samples cannot penetrate the collected output optical complex field as the complex fielddata of the input light, and adopting the complex field of a collected and transmitted to-be-reconstructed sample as the complex field data of the output light; according to a reconstruction resolution condition, adjusting the parameters of the deep learning network, and training the deep learning network; solving the three-dimensional refractive index distribution of the sample through a weight obtained through training the deep learning network, and realizing the chromatographic reconstruction of the sample. According to the invention, the chromatographic reconstruction capability low in acquisition amount and high in resolution is achieved. The resolution precision of the sample tomography reconstruction is effectively improved.

Description

technical field [0001] The present invention relates to the technical fields of computational optics, computer vision and computational photography, and in particular to a variable-angle illumination tomography method and device based on a deep learning generation network. Background technique [0002] At present, high-resolution tomographic reconstruction of microscopic samples, especially living biological samples, is a hot research issue in the fields of computational optical imaging, computer vision, and computational photography. In related tomographic techniques, since most living biological cells have the characteristics of weak intensity differences and high phase differences, phase imaging techniques are widely used for research. However, most of the existing phase tomography techniques need to collect a large amount of data, including images illuminated at different angles or focused at different depths, and the acquisition speed limits the development and applicat...

AI Technical Summary

Problems solved by technology

However, most of the existing phase tomography techniques need to collect a large amount of data, including images illuminated at different angles or focused at different depths, and the acquisition speed limits the development and application of phase tomography

[0003] Another common problem in the tomographic reconstruction of living biological samples is that the existing reconstruction techniques often have serious elongation in the direction of the optical axis, causing large errors, thus limiting the tomographic resolution in the direction of the optical axis , affecting the effect of tomographic reconstruction, it is difficult to achieve nanoscale tomography

Doing so can make the algorithm simpler and more convenient, but it will affect the tomographic effect

Images