형태와 확률 모델을 이용한 다중 차량 검출 후 추적 방법

In this paper, we present a multi-vehicle tracking-by-detection method with single mono camera. No state-of-the-art vehicle detection method can detect all the vehicles on the road and remove all false positive alarms. False negative and false positive alarms are caused by an imperfect detection algorithm. Therefore, a robust tracking-by-detection algorithm is necessary to minimize the number of false positive and false negative alarms. Many false positive alarms are removed during track initialization procedure in Pre-Track state due to their sparseness. Particle prediction can minimize the number of false negative alarms in Cur-Track and Post-Track states. The experimental results demonstrate that our multi-vehicle detection-bytracking method remarkably reduces false positive and false negative alarms when comparing with the vehicle detection algorithm.