Multisensor fusion based tracking of coalescing objects in urban environment for an autonomous vehicle navigation

This article addresses detection, association and tracking of coalescing objects in urban environment for autonomous navigation using multiple sensors. The stress is put on providing description of surroundings for the task of local navigation and collision avoidance. An Oriented Bounding Box representation is proposed for dynamic obstacles, as well as data association technique suitable for the used representation. Fusion of multiple sensors is used in order to increase the accuracy and the robustness of the tracking system. The sensors configuration consists of two laser range finders and a LIDAR.