Linear-Time Joint Probabilistic Data Association for Multiple Extended Object Tracking
Shishan Yang, Kolja Thormann, Marcus Baum · 2018
The Joint Probabilistic Data Association (JPDA) filter for multiple object tracking is based on the assumption that at most one measurement originates from a target object. However, with the development of high-resolution sensors, it is often the case that multiple spatially distributed detections are obtained from a single object. To tackle this emerging data association challenge, this paper presents a JPDA method based on the Poisson spatial measurement model for extended objects. As the constraint that one target gets at most one measurement is relaxed, the marginal association probabilities can be obtained with linear complexity in the number of measurements and targets. The proposed method is compared to a partition-based multiple extended object tracking algorithm.