Establishing a high-accuracy, dual-wavelength Mueller pulsed LiDAR testbed

Abstract

Polarization enhancement of remote sensing systems can improve sample discrimination and even resolution in ranged or imaged scenes. Assistive and autonomous machine vision systems benefit from the added data dimension of polarization in both passive and active remote sensor systems. While optimal transmit and receive polarization states have been discovered for atmospheric and oceanographic measurement scenarios, such optimal states have not yet been solidified for machine vision scenarios involving opaque samples in complex scenes, particularly when using monostatic or near-monostatic systems like lidars. This is, in part, due to a knowledge gap for material polarized response characterization, and the complex nature of full polarimetry needed to perform initial characterizing measurements before reaching optimal states. This work has four primary objectives: (1) to design a high-accuracy and precision polarimeter, (2) to develop a procedure for characterizing the monostatic polarized response of materials which is publicly available and easy to implement, (3) to develop simple calibration methods to establish accuracy and precision metrics, and (4) to pull the prior objectives together along with a pulsed lidar design to generate a dichroic Mueller pulsed lidar testbed. Drawing on literature and applying optical and mechanical tolerances, we developed an optimal polarimeter, then described the design, build, and operating procedure of that polarimeter for near-monostatic and monostatic measurements in a publication for easy replication by other groups with interest in the outcomes. Generalized calibration procedures are presented in this dissertation for public consumption, as well as the design, calibration, and validation of the primary instrument testbed. The final lidar testbed has mean Mueller coefficient errors of less than 1%, allowing it to operate as a characterization and validation instrument for the pursuit of optimal polarization states for enhanced machine vision scenarios.