Fourier Ptychographic Microscopy image aberration correction method based on multi-feature fusion residual network

Abstract Abstract
Objective: The study aims to address optical aberrations caused by sample non-uniformity or imaging system deviations in Fourier Ptychographic Microscopy (FPM), a technique enabling high-resolution microscopic imaging with a large field of view. The proposed approach involves utilizing a multi-feature fusion residual network method for aberration correction, leveraging residual networks and multi-scale convolutional kernels to extract deep residual features and enhance the ability to capture relevant image characteristics.
Approach: This method optimizes network performance using grouped convolutions and attention mechanisms without significantly increasing the parameter count. By combining Zernike modes, accurate estimation and correction of aberrations are achieved, leading to a significant improvement in imaging quality.
Main results: The research demonstrates the effective correction of aberrations using this method, enabling high-resolution image reconstruction while preserving rich texture details. Validation results based on simulated and real images showcase the method's efficacy.
Significance: By introducing the multi-feature fusion residual network method, this study addresses a significant challenge posed by aberrations in FPM, enhancing imaging quality and offering a new effective approach for high-resolution microscopic imaging. This has important implications for fields such as biomedicine that demand strict image quality standards. Keywords: Fourier Ptychographic Microscopy; Aberration Correction; Multi-Feature Fusion; Residual Network; Zernike Modes