Heterogeneous Sensor Fusion for Primary Vehicle Detection

S. Yang, B. Song, and J. Um (Korea)


Vehicle detection, Sensor fusion, Target classification, Probabilistic data association filter


In this paper, the sensor fusion algorithm is presented that the primary vehicle, i.e., the closest preceding vehicle in the same lane, is recognized more reliably by fusing radar, monocular vision, and yaw-rate sensor data. In general, most of commercial radars may often lose detection and tracking of the primary vehicle when it completely stops or goes with other preceding vehicles in the adjacent lane with similar velocity and range. In order to improve this performance degradation of radar, vehicle detection information measured by radar and vision, and path prediction predicted by a yaw-rate sensor will be combined for target classification. Then, the target classification will be cooperated with a probabilistic data association filter (PDAF) to track the primary vehicle. Finally the proposed sensor fusion algorithm will be validated using field test data on highway.

Important Links:

Go Back