Publicação

Spatial domain LiDAR

Ver documento

Detalhes bibliográficos
Resumo:This document presents the work developed on the scope of a master thesis on “Spatial domain LiDAR”. The goal of this project is to explore an alternative path to mainstream LiDARs based on time-of-flight. The alternative path consists in illuminating a scene with a dot matrix, estimating the distance to each dot via triangulation. The investigated LIDAR is validated using a custom-made ray optics simulation tool. Estimation accuracy is assessed by comparing different centroid estimation methods, investigating how relevant are setup impairments, and looking at both practical and theoretical performance limits. New solutions are proposed, namely a super-resolution method for increasing accuracy, a single-camera setup, and a method for decoupling estimation from illumination, at the same time providing robustness against interference. A preliminary experimental validation is also presented, which produced a correct point cloud of a test scenario. This work shows that the explored alternative LiDAR is viable, being an option to consider for the automotive industry.
Autores principais:Afonso, Renato Miguel
Ano:2019
País:Portugal
Tipo de documento:dissertação de mestrado
Tipo de acesso:acesso aberto
Instituição associada:Universidade de Aveiro
Idioma:inglês
Origem:RIA - Repositório Institucional da Universidade de Aveiro
Descrição
Resumo:This document presents the work developed on the scope of a master thesis on “Spatial domain LiDAR”. The goal of this project is to explore an alternative path to mainstream LiDARs based on time-of-flight. The alternative path consists in illuminating a scene with a dot matrix, estimating the distance to each dot via triangulation. The investigated LIDAR is validated using a custom-made ray optics simulation tool. Estimation accuracy is assessed by comparing different centroid estimation methods, investigating how relevant are setup impairments, and looking at both practical and theoretical performance limits. New solutions are proposed, namely a super-resolution method for increasing accuracy, a single-camera setup, and a method for decoupling estimation from illumination, at the same time providing robustness against interference. A preliminary experimental validation is also presented, which produced a correct point cloud of a test scenario. This work shows that the explored alternative LiDAR is viable, being an option to consider for the automotive industry.