Efficient colorful Fourier ptychographic microscopy reconstruction with wavelet fusion

Fourier ptychographic microscopy (FPM) is a recently developed microscope technology that overcomes the resolution limit of a low numerical aperture (NA) objective lens by employing angular varying illuminations. Combining the concepts of ptychography, synthetic aperture and phase retrieval, FPM achieves high-resolution, wide-field and quantitative phase imaging at the same time. In typical FPM systems, the angular varying illuminations are achieved with LED arrays whose positional misalignments bring significant errors in the reconstruction procedure. In previous studies, several LED array alignment methods are developed, which iteratively recover the positional misalignment parameters during the reconstruction. These methods consume additional calculations in FPM reconstruction and may not be practical in other microscopy system. In this work, we represent a preprocessing LED array alignment method by accurately localizing the brightfield area on the sample plane. By applying Particle Swarm Optimization (PSO) method and Random Sample Consensus (RANSAC) method, the global misalignment parameters can be estimated with high accuracy and speed. Both numerical simulations and actual system experiments are carried out to evaluate the effectiveness of our method and the results show that the reconstruction quality of high-resolution images is significantly improved by using our method.

Wavelet multi-resolution color fusion of a low-resolution image and a high-resolution image.

Reconstruction performance of three methods on human blood smear images.