Sensor fusion for vehicle tracking with camera and radar sensor

In recent years, as demands for the vehicle safety and autonomous driving of the automobile industry have increased, it becomes important to more accurately recognize the position and velocity of surrounding vehicles. In this paper, heuristic fusion with adaptive gating and track to track fusion are applied to track fusion of camera and radar sensor for forward vehicle tracking system and the two algorithms are compared. To compare the two algorithms, simulation was carried out in 10 scenarios and the accuracy of sensor fusion results was measured with optimal subpattern assignment (OSPA) metric. The results of this metric are compared to show that the track to track fusion is superior to the adaptive gating for the target estimation.