Addressing Track Coalescence in Sequential K-Best Multiple Hypothesis Tracking
Ryan D. Palkki, Aaron D. Lanterman, William Dale Blair · 2006
The sequential K -best multiple hypothesis tracker is implemented for a single-target, single-sensor scenario. The K-best data association approach is compared to the probabilistic data association technique to determine under what conditions it is generally the preferred method. A fundamental problem is observed in which the tracks coalesce. Several methods to prevent coalescence are presented and compared