| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 6.79 MB | Adobe PDF |
Authors
Advisor(s)
Abstract(s)
Quando um sensor LiDAR ´e montado numa plataforma móvel, seja esta terrestre ou aérea, ´e necessário determinar a posição absoluta e a orientação do sensor para utilizar os dados obtidos, seja para modelação 3D, calculo de áreas de zonas de interesse ou, como na maioria das vezes na área da robótica, para mapeamento 3D. Os veículos responsáveis por transportar estes sensores são, normalmente capacitados de sistemas de GNSS que fornecem informações geográficas e de INS que registam a orientação dos veículos. Esta dissertação propõe endere¸car o problema de estimar a posição e orientação relativa do sensor LiDAR em relação a um UAV em diferentes cenários. Nesse sentido, pretende-se que a abordagem tomada seja capaz de detetar planos, em múltiplas orientações, numa nuvem de pontos geo-referenciada, extrair os parâmetros que os representam e utilizá-los para estimar os parâmetros extrínsecos do sensor.
When a LiDAR sensor is mounted on a mobile vehicle, been aerial or terrestrial, itis necessary to determine the absolute position and orientation of the sensor to use the obtained data, either for 3D modeling, measurement of interest zones or, as usually on robotics field, for 3D mapping. Vehicles that have to transport those sensors are also endowed with GNSS modules that provide geographical information and INS modules that record the orientation of those vehicles. This thesis proposes to address the estimation problem of the relative position and orientation between the LiDAR sensor and the UAV robot in multiple scenarios. Therefore, it is intended to the took approach to be able to detect planes in multiple orientations on a georeferenced pointcloud, extract the parameters that represent them and use them to estimate the extrinsic parameters of the sensor.
When a LiDAR sensor is mounted on a mobile vehicle, been aerial or terrestrial, itis necessary to determine the absolute position and orientation of the sensor to use the obtained data, either for 3D modeling, measurement of interest zones or, as usually on robotics field, for 3D mapping. Vehicles that have to transport those sensors are also endowed with GNSS modules that provide geographical information and INS modules that record the orientation of those vehicles. This thesis proposes to address the estimation problem of the relative position and orientation between the LiDAR sensor and the UAV robot in multiple scenarios. Therefore, it is intended to the took approach to be able to detect planes in multiple orientations on a georeferenced pointcloud, extract the parameters that represent them and use them to estimate the extrinsic parameters of the sensor.
Description
Keywords
Nuvem de Pontos LiDAR Parâmetros Extrínsecos Deteção de Planos PointCloud LiDAR Extrinsic Parameters Plane Detection