Halo-effect-free white light phase imaging method and system based on deep learning

Abstract

The invention discloses a halo-effect-free white light phase imaging method and system based on deep learning; the system comprises a white light diffraction phase imaging module, a data processing module, a deep neural network model building module and a halo-effect-free white light phase imaging module, wherein the white light diffraction phase imaging module is used for collecting a standard polystyrene microsphere phase image of each biological living cell; the data processing module is used for carrying out regularity iterative calculation on the input data to obtain a corresponding halo-effect-free phase image, and carrying out data augmentation operation on the image, so that a data set is expanded, and the data set is divided into a training set, a verification set and a test set at 7: 2: 1; the deep neural network model building module is used for building a deep neural network model and selecting the optimal model; and the halo-effect-free white light phase imaging module isused for obtaining standard polystyrene microsphere phase images with different sizes through the white light diffraction phase imaging module, and realizing the halo-effect-free white light phase imaging through calculation based on the derived optimal model. The method can eliminate the phase image halo effect obtained by the white light diffraction phase imaging system, and has the advantages of being high in instantaneity and high in accuracy.

Description

technical field [0001] The present invention relates to the technical field of phase imaging, in particular to the white light phase imaging technology without halo effect based on deep learning. Background technique [0002] White light diffraction phase microscopy has broad application prospects in the fields of materials and biological cell science. However, white light diffraction phase microscopy is illuminated by an extended white light source. The illumination light wave cannot be a complete plane light wave, and its spatial coherence area is generally Far smaller than the measurement field of view, so that the halo effect that depends on the structure of the object appears in the phase image. In fact, the phase imaging technology using white light all has halo effect. [0003] At present, there are two main types of methods to solve the influence of the halo effect. One is to improve the spatial coherence of the lighting source. Generally, the spatial coherence are...

AI Technical Summary

Problems solved by technology

One is to improve the spatial coherence of the lighting source. Generally, the spatial coherence area of ​​the lighting source is increased by reducing the numerical aperture of the lighting source. However, this method will reduce the luminous flux and increase the exposure time, which is not conducive to the realization of dynamic measurement.

Another type of method is to combine the numerical iterative calculation of the hardware parameters of the imaging system. This method can accurately correct the halo effect of the phase image, but the method of iterative calculation is time-consuming and depends on the system hardware parameters.

The other is a direct digital image processing method. Through the three-way Hilbert transform filtering method, the halo effect on the edge can be better eliminated, but the deviation between the remaining phase value and the accurate value caused by the halo effect cannot be corrected.

Images