Underwater image-based 3d reconstruction with quality estimation
- Autores: Klemen Istenič
- Directores de la Tesis: Nuno Gracias (dir. tes.), Rafael Garcia Campos (codir. tes.)
- Lectura: En la Universitat de Girona ( España ) en 2021
- Idioma: español
- Tribunal Calificador de la Tesis: Fabio Bruno (presid.), Josep Forest Collado (secret.), Dimitris Skarlatos (voc.)
- Programa de doctorado: Programa de Doctorado en Tecnología por la Universidad de Girona
- Materias:
- Enlaces
- Tesis en acceso abierto en: TDX
- Resumen
This thesis addresses the development of resources for accurate scaling and uncertainty estimation of image-based 3D models for scientific purposes based on data acquired with monocular or unsynchronized camera systems in difficult-to-access GPS-denied (underwater) environments.
The developed 3D reconstruction framework allows the creation of textured 3D models based on optical and navigation data and is independent of a specific platform, camera or mission. The dissertation presents two new methods for automatically scaling of SfM-based 3D models using laser scalers. Both were used to perform an in-depth scale error analysis of large-scale models of deep-sea underwater environments to determine the advantages and limitations of image-based 3D reconstruction strategies.
In addition, a novel SfM-based system is proposed to demonstrate the feasibility of producing a globally consistent reconstruction with its uncertainty while the robot is still in the water or shortly after.
