Tracking of vehicles on nearside lanes using multiple radar sensors

Tracking overtaking vehicles is an integral aspect of Lane Change Assistant Systems. For this, in this work, a set of six radar sensors is mounted on a test vehicle. As the sensors generate spatially extended measurements from different parts of the vehicles, the estimation of position and velocity of the surrounding vehicles leads to an extended target tracking problem. This is especially the case when a vehicle is located on a nearside lane. A method for tracking these targets is proposed, under the assumption that a target vehicle comprises a random number of reflection centers located on a rectangular structure. To estimate the existence of a reflector in combination with the vehicle target state, the Joint Integrated Probabilistic Data Association Filter (JIPDA) is applied.