Detail výsledku
HDR and 3D Occupancy map implementation on Embedded HW
Nosko Svetozár, Ing., Ph.D., UPGM (FIT)
Beran Vítězslav, doc. Ing., Ph.D., UPGM (FIT)
Zemčík Pavel, prof. Dr. Ing., dr. h. c., UPGM (FIT)
This technical report describes the implementation of selected methods for sensory data processing on embedded HW. Outdoor 3D environment modeling provides volumetric space information that is important in solving some robotic tasks, including unmanned areal vehicles with remote control but with limited visual connection. To this end, accurate 3D occupancy maps and reliable flight control are key to the task. This report represents the generation of a 3D occupancy map from a 3D point cloud and its realisation on embedded HW. In addition, the ZED2 stereoscopic camera is set up and tested to generate a 3D point clouds. Finally, we expanded the HDR camera (High Dynamic Range) with a new HDH deghosting algorithm, which enabled the integration of HDR shooting directly on the drone. This algorithm reduces the adverse effects that can occur during the flight.
@misc{BUT168686,
author="Alfredo {Chavez Plascencia} and Svetozár {Nosko} and Vítězslav {Beran} and Pavel {Zemčík}",
title="HDR and 3D Occupancy map implementation on Embedded HW",
year="2020",
pages="1--27",
address="Brno",
url="https://www.fit.vut.cz/research/publication/12435/"
}