Imagine of having an autonomous agent (drone, robot, car, ..) that wants to navigate inside an unknown environment. The first question that it needs to answer for accomplish such task is: where Am I? Where are the objects that are surrounding me? The SLAM algorithm can answer to both questions simultaneously, in an on-line manner. This thesis focus on the implementation of a monocular SLAM algorithm on the UAV framework, where the classical obtained sparsity map is densified by means of a Convolutional Neural Network, properly scaled through 2D lidar measurements.
Imagine of having an autonomous agent (drone, robot, car, ..) that wants to navigate inside an unknown environment. The first question that it needs to answer for accomplish such task is: where Am I? Where are the objects that are surrounding me? The SLAM algorithm can answer to both questions simultaneously, in an on-line manner. This thesis focus on the implementation of a monocular SLAM algorithm on the UAV framework, where the classical obtained sparsity map is densified by means of a Convolutional Neural Network, properly scaled through 2D lidar measurements.
Lidar-based scale recovery dense SLAM for UAV navigation
ANDREOLI, JACOPO
2021/2022
Abstract
Imagine of having an autonomous agent (drone, robot, car, ..) that wants to navigate inside an unknown environment. The first question that it needs to answer for accomplish such task is: where Am I? Where are the objects that are surrounding me? The SLAM algorithm can answer to both questions simultaneously, in an on-line manner. This thesis focus on the implementation of a monocular SLAM algorithm on the UAV framework, where the classical obtained sparsity map is densified by means of a Convolutional Neural Network, properly scaled through 2D lidar measurements.File | Dimensione | Formato | |
---|---|---|---|
Andreoli_Jacopo.pdf
accesso aperto
Dimensione
10.44 MB
Formato
Adobe PDF
|
10.44 MB | Adobe PDF | Visualizza/Apri |
The text of this website © Università degli studi di Padova. Full Text are published under a non-exclusive license. Metadata are under a CC0 License
https://hdl.handle.net/20.500.12608/36256