Efficient Colorful Fourier Ptychographic Microscopy Reconstruction With Wavelet Fusion

Abstract

Fourier ptychographic microscopy (FPM) is a recently developed microscope that offers wide field, high resolution, and quantitative phase imaging. Combining the concepts of ptychography, synthetic aperture, and phase retrieval, FPM overcomes the space-bandwidth-product (SBP) limit of the optical system. In FPM, a LED array is used as the illumination module and specimen images illuminated with angular varying illuminations are captured. These images are synthesized to recover the high SBP complex field of the specimen. There have been many improvements of FPM since its appearance and FPM is of great potential in the field of hematology and pathology. However, low efficiency in capturing data, especially in the situation of capturing all three color channels, limits the application of FPM. Although spectral multiplex strategy for FPM is developed, the reconstruction quality and speed is decreased for exchanging the capture efficiency. On the other hand, the reconstruction quality of FPM is significantly degraded by the imaging noise in the dark-field images because the high-frequency information in dark-field images is of low energy and contaminated by noise. In this paper, we propose an efficient colorful FPM reconstruction method using multi-resolution wavelet color fusion. We also propose an adaptive denoising method by analyzing the noise information of the dark-frame. Both simulation and experiment results are carried out to validate our methods. Results demonstrate that the imaging noise is suppressed and the colorful reconstruction is of high efficiency and quality.