Abstract

Quantitative phase image (QPI) is a popular microscopy technique for studying cell morphology. Recently, Fourier ptychographic microscopy (FPM) has emerged as a low-cost computational microscopy technique for forming high-resolution wide-field QPI images by taking multiple images from different illumination angles. However, the applicability of FPM to dynamic imaging is limited by its high data requirement. In this thesis, we propose new methods for highly compressive FPM imaging using a data-adaptive sparse coding and an online plug-and-play (PnP) method with non-local priors based on the fast iterative shrinkage/threshold algorithm (FISTA). We validate the proposed method on both simulated and experimental data and show that our method is capable of reconstructing images under a significantly lower data rate.

Degree

Master of Science (MS)

Author's Department

Electrical & Systems Engineering

Author's School

McKelvey School of Engineering

Document Type

Thesis

Date of Award

Spring 2019

Language

English (en)

Share

COinS