Multi-target data association approach for vehicle tracking in road situation
Hongshe Dang, Chongzhao Han · 2004
Data association was an important content in Multi-target tracking. Typical algorithms to deal with such problems are the joint probabilities data association filter (JPDAF) proposed by Bar-Shalom and his team. The basis of JPDAF is the calculus of the joint probabilities between the measurements and the tracks. The algorithm assigns weights for reasonable measurements and uses a weighted centroid of those measurements to update the track. Based on the evidence theory and fuzzy mathematics, a new data association method between the tracks and the measurements is proposed in the paper. The mass function is determined by using fuzzy mathematics, and a belief matrix is produced according to the evidence combination rule, then the decision is made by means of the maximum belief. The Monte Carlo simulation results indicate that the new method has a good association capacity. Compared with cheap JPDAF method, the new method has improved the track accuracy.