HAL will be down for maintenance from Friday, June 10 at 4pm through Monday, June 13 at 9am. More information
Skip to Main content Skip to Navigation
Master Thesis

Analyse des données LIDAR issues de la base nuScenes. Qualification (semi-)automatique de l’analyse des interactions entre véhicules

Abstract : Nowadays, autonomous vehicles represent a potential new source of information in the understanding and study of traffic, whether in real time or after the fact. This data, which is very rich in interaction details and increasingly numerous, is nevertheless little exploited in the field of road traffic. For the moment, they are mainly used in the problem of the perception of the environment by the autonomous vehicle. The work made during this study consists in determining the information an autonomous vehicle database can provide based on the nuScenes database. NuScenes was created based on experiments with autonomous vehicles in real traffic conditions. This work aims, firstly, to characterize the interactions between two vehicles as rendered by the passage of an autonomous vehicle and, secondly, to group the observed interactions in an unsupervised way and to compare them to manually identified categories. The underlying objective is, in the long term, to be able to automatically identify the interactions between two vehicles referring to a vehiclefollower type of behaviour and to quantify the differences in behaviour between human and automated driving.
Document type :
Master Thesis
Complete list of metadata

Contributor : Bibliothèque Entpe Connect in order to contact the contributor
Submitted on : Thursday, March 17, 2022 - 10:53:39 AM
Last modification on : Wednesday, April 13, 2022 - 3:27:31 AM


Files produced by the author(s)




  • HAL Id : dumas-03611480, version 1



Tiffany Darini. Analyse des données LIDAR issues de la base nuScenes. Qualification (semi-)automatique de l’analyse des interactions entre véhicules. Sciences de l'ingénieur [physics]. 2021. ⟨dumas-03611480⟩



Record views


Files downloads