When
Oct. 23, 2026, 3:30 – 6:30 p.m.
Where
Meinel 821
Title:
Optical Encoding and Computational Reconstruction for Imaging and Sensing
Abstract:
Modern imaging systems often seek information that cannot be measured directly by conventional detectors, including three-dimensional geometry, surface reflectance, and the amplitude and phase of optical fields. This dissertation explores how optical encoding and computational reconstruction can be jointly designed to recover such information, focusing on three complementary encoding domains: illumination, pupil/Fourier space, and image-data representation.
First, randomized RGB structured illumination enables joint recovery of 3D geometry and color from a single scene exposure using a calibrated, differentiable imaging model. Second, overlapping pupil measurements acquired through metasurface and camera-array architectures enable complex-wavefront reconstruction and digital refocusing. Third, integrated photonic encoding provides optical dimensionality reduction of image data, combined with computational decoding. Together, these studies establish a framework for co-designing optical measurements, calibration, and reconstruction, highlighting how information can be acquired, constrained, and efficiently represented for future computational imaging and sensing systems.
Committee:
Dr. David Brady (Chair)
Dr. Jose Sasian
Dr. Weimin Zhou
Dr. John Koshel
Please email Xiao Wang at xwang3@arizona.edu or graduate student advisor Jini Kandyil (jini@optics.arizona.edu) for the Zoom link.