Impact of calibration in a LiDAR based on stereoscopic vision
Autor(a) principal: | |
---|---|
Data de Publicação: | 2022 |
Tipo de documento: | Dissertação |
Idioma: | eng |
Título da fonte: | Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
Texto Completo: | http://hdl.handle.net/10773/38422 |
Resumo: | Every year 1.3 million people die due to road accidents. Given that the main culprit is human error, autonomous driving is the path to avert and prevent these numbers. An autonomous vehicle must be able to perceive its surroundings, therefore requiring vision sensors. Of the many kinds of vision sensors available, the three main automotive vision sensors are cameras, RADAR and LiDAR. LiDARs have the unique capability of capturing a high-resolution point cloud, thus enabling 3D object detection. However, current LiDAR technology is still immature and expensive, which makes it unattractive to the automotive market. We propose an alternative LiDAR concept – the LiDART – that is able to generate a point cloud simply resorting to stereoscopic vision and dot projection. LiDART takes advantage of mass-produced components such as a dot pattern projector and a stereoscopic camera rig, thus inherently overcoming problems in cost and maturity. Nonetheless, LiDART has four key challenges: noise, correspondence, centroiding and calibration. This thesis focuses on the calibration aspects of LiDART and aims to investigate the systematic error introduced by standard calibration techniques. In this work, the quality of stereoscopic calibration was assessed both experimentally and numerically. The experimental validation consisted in assembling a prototype and calibrating it using standard calibration techniques for stereoscopic vision. Calibration quality was assessed by estimating the distance to a target. As for numerical assessment, a simulation tool was developed to cross-validate most experimental results. The obtained results show that standard calibration techniques result in a considerable systematic error, reaching 30% of the correct distance. Nonetheless, the estimated error depends monotonically on distance. Consequently, the systematic error can be significantly reduced if better calibration methods, specifically designed for the application at hand, are used in the future. |
id |
RCAP_d33446eef9766c1e61330b4757e8adbd |
---|---|
oai_identifier_str |
oai:ria.ua.pt:10773/38422 |
network_acronym_str |
RCAP |
network_name_str |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
repository_id_str |
7160 |
spelling |
Impact of calibration in a LiDAR based on stereoscopic visionAutonomous vehiclesComputer visionLiDARStereo imagingCalibrationEvery year 1.3 million people die due to road accidents. Given that the main culprit is human error, autonomous driving is the path to avert and prevent these numbers. An autonomous vehicle must be able to perceive its surroundings, therefore requiring vision sensors. Of the many kinds of vision sensors available, the three main automotive vision sensors are cameras, RADAR and LiDAR. LiDARs have the unique capability of capturing a high-resolution point cloud, thus enabling 3D object detection. However, current LiDAR technology is still immature and expensive, which makes it unattractive to the automotive market. We propose an alternative LiDAR concept – the LiDART – that is able to generate a point cloud simply resorting to stereoscopic vision and dot projection. LiDART takes advantage of mass-produced components such as a dot pattern projector and a stereoscopic camera rig, thus inherently overcoming problems in cost and maturity. Nonetheless, LiDART has four key challenges: noise, correspondence, centroiding and calibration. This thesis focuses on the calibration aspects of LiDART and aims to investigate the systematic error introduced by standard calibration techniques. In this work, the quality of stereoscopic calibration was assessed both experimentally and numerically. The experimental validation consisted in assembling a prototype and calibrating it using standard calibration techniques for stereoscopic vision. Calibration quality was assessed by estimating the distance to a target. As for numerical assessment, a simulation tool was developed to cross-validate most experimental results. The obtained results show that standard calibration techniques result in a considerable systematic error, reaching 30% of the correct distance. Nonetheless, the estimated error depends monotonically on distance. Consequently, the systematic error can be significantly reduced if better calibration methods, specifically designed for the application at hand, are used in the future.Todos os anos 1.3 milhões de pessoas perdem a vida devido a acidentes de viação. Dado que a principal razão por detrás destes trágicos números é o erro humano, o caminho para prevenir perder tantas vidas passa pela condução autónoma. Um veículo autónomo deve ser capaz de observar o cenário envolvente. Para tal, são necessários sensores de visão. Dos vários sensores de visão disponiveis no mercado, os três principais sensores de visão automotivos são a câmara, o RADAR e o Li- DAR. O LiDAR tem a capacidade única de capturar uma nuvem de pontos com alta resolução, permitindo assim deteção de objetos em 3D. Contudo, a tecnologia por detrás de um LiDAR é atualmente dispendiosa e imatura, o que tem dificultado a adoção por parte de fabricantes de automóveis. Este trabalho propõe um conceito de LiDAR alternativo – o LiDART – capaz de gerar uma nuvem de pontos recorrendo simplesmente a visão estereoscópica e à projeção de pontos. O LiDART tem a vantagem de se basear em componentes produzidos em massa, tais como um projector de pontos e uma câmara estereoscópica, ultrapassando assim os problemas de custo e maturidade. Não obstante, o LiDART tem quatro desafios principais: ruído, correspondência, estimação de centróide e calibração. Esta tese foca-se nas características de calibração do LiDART, tendo como objectivo investigar o erro sistemático introduzido por técnicas de calibração comuns. A qualidade da calibração foi avaliada experimentalmente e numericamente. A validação experimental consistiu em montar um protótipo e calibrá-lo de várias maneiras. A qualidade da calibração foi então avaliada através da estimação da distância a um alvo. Relativamente à parte numérica, desenvolveu-se uma ferramenta de simulação para validar grande parte dos resultados experimentais. Os resultados obtidos mostram que técnicas de calibração comuns resultam num erro sistemático considerável, chegando a 30% da distância correta. Porém, o erro de estimação varia monotonicamente com a distância. Consequentemente, o erro sistemático pode ser reduzido significativamente se melhores métodos de calibração, especialmente pensados para a aplicação em questão, forem aplicados no futuro.2023-07-07T13:49:05Z2022-12-13T00:00:00Z2022-12-13info:eu-repo/semantics/publishedVersioninfo:eu-repo/semantics/masterThesisapplication/pdfhttp://hdl.handle.net/10773/38422engAlegria, João Jesus de Sousainfo:eu-repo/semantics/openAccessreponame:Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)instname:Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informaçãoinstacron:RCAAP2024-02-22T12:14:16Zoai:ria.ua.pt:10773/38422Portal AgregadorONGhttps://www.rcaap.pt/oai/openaireopendoar:71602024-03-20T03:08:35.419176Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) - Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informaçãofalse |
dc.title.none.fl_str_mv |
Impact of calibration in a LiDAR based on stereoscopic vision |
title |
Impact of calibration in a LiDAR based on stereoscopic vision |
spellingShingle |
Impact of calibration in a LiDAR based on stereoscopic vision Alegria, João Jesus de Sousa Autonomous vehicles Computer vision LiDAR Stereo imaging Calibration |
title_short |
Impact of calibration in a LiDAR based on stereoscopic vision |
title_full |
Impact of calibration in a LiDAR based on stereoscopic vision |
title_fullStr |
Impact of calibration in a LiDAR based on stereoscopic vision |
title_full_unstemmed |
Impact of calibration in a LiDAR based on stereoscopic vision |
title_sort |
Impact of calibration in a LiDAR based on stereoscopic vision |
author |
Alegria, João Jesus de Sousa |
author_facet |
Alegria, João Jesus de Sousa |
author_role |
author |
dc.contributor.author.fl_str_mv |
Alegria, João Jesus de Sousa |
dc.subject.por.fl_str_mv |
Autonomous vehicles Computer vision LiDAR Stereo imaging Calibration |
topic |
Autonomous vehicles Computer vision LiDAR Stereo imaging Calibration |
description |
Every year 1.3 million people die due to road accidents. Given that the main culprit is human error, autonomous driving is the path to avert and prevent these numbers. An autonomous vehicle must be able to perceive its surroundings, therefore requiring vision sensors. Of the many kinds of vision sensors available, the three main automotive vision sensors are cameras, RADAR and LiDAR. LiDARs have the unique capability of capturing a high-resolution point cloud, thus enabling 3D object detection. However, current LiDAR technology is still immature and expensive, which makes it unattractive to the automotive market. We propose an alternative LiDAR concept – the LiDART – that is able to generate a point cloud simply resorting to stereoscopic vision and dot projection. LiDART takes advantage of mass-produced components such as a dot pattern projector and a stereoscopic camera rig, thus inherently overcoming problems in cost and maturity. Nonetheless, LiDART has four key challenges: noise, correspondence, centroiding and calibration. This thesis focuses on the calibration aspects of LiDART and aims to investigate the systematic error introduced by standard calibration techniques. In this work, the quality of stereoscopic calibration was assessed both experimentally and numerically. The experimental validation consisted in assembling a prototype and calibrating it using standard calibration techniques for stereoscopic vision. Calibration quality was assessed by estimating the distance to a target. As for numerical assessment, a simulation tool was developed to cross-validate most experimental results. The obtained results show that standard calibration techniques result in a considerable systematic error, reaching 30% of the correct distance. Nonetheless, the estimated error depends monotonically on distance. Consequently, the systematic error can be significantly reduced if better calibration methods, specifically designed for the application at hand, are used in the future. |
publishDate |
2022 |
dc.date.none.fl_str_mv |
2022-12-13T00:00:00Z 2022-12-13 2023-07-07T13:49:05Z |
dc.type.status.fl_str_mv |
info:eu-repo/semantics/publishedVersion |
dc.type.driver.fl_str_mv |
info:eu-repo/semantics/masterThesis |
format |
masterThesis |
status_str |
publishedVersion |
dc.identifier.uri.fl_str_mv |
http://hdl.handle.net/10773/38422 |
url |
http://hdl.handle.net/10773/38422 |
dc.language.iso.fl_str_mv |
eng |
language |
eng |
dc.rights.driver.fl_str_mv |
info:eu-repo/semantics/openAccess |
eu_rights_str_mv |
openAccess |
dc.format.none.fl_str_mv |
application/pdf |
dc.source.none.fl_str_mv |
reponame:Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) instname:Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informação instacron:RCAAP |
instname_str |
Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informação |
instacron_str |
RCAAP |
institution |
RCAAP |
reponame_str |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
collection |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) |
repository.name.fl_str_mv |
Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos) - Agência para a Sociedade do Conhecimento (UMIC) - FCT - Sociedade da Informação |
repository.mail.fl_str_mv |
|
_version_ |
1799137738258120704 |