In the past few years, autonomous vehicles have been the subject of innovation and technology amidst the challenges and difficult situations surrounding the urban driving and infrastructure. The recent developments in the sensors and embedded systems made way to do a deeper analysis of the cost and needs of AV. This study aims to make use of these new technologies and sensor to have a better recognition model of the relevant road actors surrounding autonomous vehicles. The results of this study made it possible to use depth and RGB data to recognize obstacles in an urban road driving scenario. The mean average precision of the model across all labels shows acceptable results running in ten frames per second setting. The model was deployed in an autonomous golf picker buggy using Robotic Operating System.
Obstacle Recognition Using Depth Estimation and RGB Data for Autonomous Vehicle
Lect. Notes in Networks, Syst.
Proceedings of the Future Technologies Conference ; 2022 ; Vancouver, BC, Canada October 20, 2022 - October 21, 2022
Proceedings of the Future Technologies Conference (FTC) 2022, Volume 3 ; Kapitel : 32 ; 478-484
14.10.2022
7 pages
Aufsatz/Kapitel (Buch)
Elektronische Ressource
Englisch
Obstacle recognition method based on autonomous vehicle and autonomous vehicle
Europäisches Patentamt | 2022
|APPARATUS AND METHOD FOR IMPROVING OBSTACLE RECOGNITION PERFORMANCE OF AUTONOMOUS VEHICLE
Europäisches Patentamt | 2025
APPARATUS AND METHOD FOR EVALUATING OBSTACLE RECOGNITION PERFORMANCE OF AUTONOMOUS VEHICLE
Europäisches Patentamt | 2023
APPARATUS AND METHOD FOR EVALUATING OBSTACLE RECOGNITION PERFORMANCE OF AUTONOMOUS VEHICLE
Europäisches Patentamt | 2023
|