Princeton University

School of Engineering & Applied Science

Computational Methods for Microfluidic Microscopy and Phase-space Imaging

Nicolas Pégard
125 Sherrerd Hall
Tuesday, August 5, 2014 - 9:00am to 10:30am

In this thesis, we present general methods for improving the performance of optical systems. By  combining all the design constraints into a single optimization problem, we provide task specific solutions for a broad range of optical systems and show examples in optical data storage, holography, microscopy and even photovoltaic devices.

We have implemented tools such as Fractional Fourier Transforms to represent the propagation of optical signals, and we have observed that holographic storage of out-of-focus images is more robust to losses of information. This result led to the conception of a new type of microscope capable of optical sectioning.

In a second part, we introduce microfluidic flow transport as an extra degree of freedom for imaging and show how putting samples in motion not only improves image quality, but also provides extra resolution, 3D capabilities, and phase information.

Finally, we show that our methods not only benefit imaging sciences, but also solar power research. We use light-field representations to understand the optical properties of surface folds on layered organic photovoltaic devices and provide methods to find the surface pattern that bests converts photons into electricity.