Title

Polarimetric Imaging

Date of Award

5-6-2023

Degree Name

Ph.D. in Electrical and Computer Engineering

Department

Department of Electrical and Computer Engineering

Advisor/Chair

Keigo Hirakawa

Abstract

Focal plane arrays in polarimetric imaging require demodulation of the spatially modulated analyzers in order to interpolate or recover the full frame analyzer images. These micro-polarizer array (MPA) sensors are intended for real time video use, and therefore are subject to noise due to short integration time. To address these limitations together would require a joint demosaicking and denoising algorithm. We first propose a demosaicking framework to improve existing demosaicking methods, and a division-of-time Stokes simplex polarimetric imaging denoising algorithm to individually address the issues. We then develop a joint demosaicking and denoising algorithm that combines the Stokes simplex analysis with the wavelet-based denoising. The experimental results prove superior reconstruction for the joint demosaicking and denoising algorithm when compared to state-of-the-art denoising and demosaicking algorithms applied separately to the MPA.

Keywords

Electrical Engineering, Polarimetric Imaging, Demosaicking, Denoising

Rights Statement

Copyright 2023, author

Link to Full Text
COinS