3D Reconstruction of Civil Infrastructures from UAV Lidar point clouds

• Autor(a) principal: Leonardo Gomes Ribeiro
• Data de Publicação: 2021
• Tipo de documento: Dissertação
• Idioma: eng
• Título da fonte: Repositório Científico de Acesso Aberto de Portugal (Repositórios Cientìficos)
• Texto Completo: https://hdl.handle.net/10216/135354
• topic: Engenharia electrotécnica, electrónica e informática; Electrical engineering, Electronic engineering, Information engineering

Nowadays, infrastructures for transportation, communication, energy, industrial production and social purpose are presented as pillars of society, being essential for its proper functioning. Coupled with this great importance within society, there is a need to ensure the safety and durability of these assets. Thus, reliable techniques should be used to assess their condition. With technological advances and the development of new methods of data acquisition, some tasks related to civil construction, currently performed by human beings, such as inspection and quality control, become inefficient due to their danger and cost. In this context, 3D reconstruction of infrastructures appears as a possible solution, presenting itself as a first step for the monitoring of infrastructures, as well as a tool for semi or completely automated inspection processes. For the development of this thesis, a Lidar sensor coupled to a UAV was used. With this equipment, it became possible to autonomously fly over real infrastructures, extracting data from all its surfaces, regardless of the difficulties that could arise to reach such regions from the ground. The data is extracted in the form of point clouds with respective intensities, filtered, and used in reconstruction and texturing algorithms, culminating in a virtual and three-dimensional representation of the target infrastructure. With these representations, it is possible to evaluate the evolution of the infrastructure during its construction or repair, as well as to evaluate the temporal evolution of certain defects present in the construction, by comparing models for the same scenario obtained from data extracted on different occasions. This approach allows the process of monitoring infrastructures to be carried out more efficiently, with lower costs and ensuring the safety of workers.