Back to results
University of Nevada - Reno
Loosely coupled LiDAR-Visual/Thermal-Inertial Odometry and Mapping
Abstract
dc:description.abstractThis thesis presents a loosely coupled LiDAR-Visual/Thermal-Inertial odometry and mapping method that uses factor graphs. Our approach jointly optimizes relative pose constraints provided by a LiDAR scan-to-scan alignment method and a Visual/Thermal-Inertial method with preintegrated IMU constraints. An optimized relative pose prior is provided to a LiDAR scan-to-map alignment method to finally output the odometry of the system as well as globally registered pointclouds. A set of evaluation studies is presented showing the outperformance of our approach against the LiDAR only odometry and mapping method in a tunnel environment and the field-verification in an autonomous mission in an underground mine.
Degree
thesis:*- Level thesis:degree_level
- Master's Degree
- Year dc:date.issued
- 2020
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Khedekar, Nikhil Vijay
- Advisor dc:contributor.advisor
-
- Alexis, Konstantinos
- Committee members dc:contributor.committeemember
-
- Scaramuzza, Davide
- Papachristos, Christos
- Panorska, Anna
Rights
dc:rights- Statement dc:rights
-
- Creative Commons Attribution 4.0 United States
Identifiers
dc:identifier.*- Handle dc:identifier.uri
- http://hdl.handle.net/11714/7690
- OAI identifier oai:identifier
- oai:scholarwolf.unr.edu:11714/7690