Occupancy Zone기반의 Convex Interpolation에 의한 객체 차량 추적

Abstract

For autonomous vehicles equipped with advanced active safety systems, it is essential to track the behavior of the target vehicle. Radar is suitable as an automotive sensor, but it has measurement uncertainty due to the operating mechanism and reflection points. In this paper, we propose a method to reduce the tracking error of the radar system. We divide the zones uniformly and use relative location information to identify surrounding vehicles and then analyze the error characteristics of each zone that is uniformly divided. We then used the convex interpolation method to effectively track the object. The resulting estimate is more accurate than the radar measurement.