Point Cloud Registration with Limited Computational Resources using ToF Cameras

• Pablo Bernal-Polo ^[1] ; Humberto Martínez-Barberá ^[1] ; David Herrero-Pérez ^[2]
  1. [1] Universidad de Murcia

     Universidad de Murcia

     Murcia, España

     
  2. [2] Universidad Politécnica de Cartagena

     Universidad Politécnica de Cartagena

     Cartagena, España

     
• Localización: Proceedings of the XXV International Workshop on Physical Agents / David Herrero Pérez (ed. lit.), Humberto Martínez Barberá (ed. lit.), Pablo Bernal Polo (ed. lit.), Nieves Pavón Pulido (ed. lit.), 2025, ISBN 978-84-10327-06-1, págs. 150-164
• Idioma: inglés
• Enlaces
  • Texto Completo Libro
• Resumen

  • This work stems from a project aimed at designing an Unmanned Aerial Vehicle (UAV) for sewer inspection. Localization is keyfor successful operation in such scenario. This paper explores the topic of point cloud registration for its execution on a platform with limited computational resources, such as the embedded Raspberry Pi system.The presented work evaluates the feasibility of using 3D Time-of-Flight (ToF) camera sensors to achieve localization in GNSS-denied environments. Point cloud registration algorithms specifically tailored for the application at hand are implemented, and their performance and feasibility are evaluated.