A hybrid genetic algorithm for multi-hypothesis tracking

Multi-hypothesis tracking is the process of tracking the paths traversed by multiple targets. The assignment of target detections in each scan to the existing tracks or forming new tracks is called the Data Association Problem. Since the tracking is continuous and there are multiple targets, the data association problem becomes a Combinatorial Problem. As it is impractical to find an optimal solution for combinatorial problems, Heuristic approaches are encouraged. These Approaches tend to produce sub-optimal solutions in a fixed amount of time. Recent researches show that Genetic Algorithms (GA) can be applied to combinatorial problems and the results are promising. In this paper, a near-optimal solution for the Multi-hypothesis tracking problem using a Hybrid Genetic Algorithm (HGA) has been discussed.