Section: New Results

2D Laser Based Road Obstacle Classification for Road Safety Improvement

Participants : Pierre Merdrignac, Evangeline Pollard, Fawzi Nashashibi.

Vehicle and pedestrian collisions often result in fatality to the vulnerable road users (VRU), indicating a strong need to protect such persons. Laser sensors have been extensively used for moving obstacles detection and tracking. Laser impacts are produced by reflection on these obstacles which suggests an information is available to recognize multiple road obstacles classes (pedestrian, cyclists, vehicles,...). We introduce a new system to address this problem that is divided in three parts: definition of geometric features describing road obstacles, multi-class object classification from an adaboost trained classifier and Bayesian estimation of the obstacle class. This approach benefits from consecutive observations of a single obstacle to estimate its class more precisely. We tested our system on some laser sequences and showed that it can estimate the class of some road obstacles around the vehicle with an accuracy of 87.4%. The vehicle class is determined with more than 97% of success. However, the main source of confusion is for static obstacles (posts and trees) for which 15% are classified as pedestrians. More detail can be fund in [36] , [16] .