A density-based recursive RANSAC algorithm for unmanned aerial vehicle multi-target tracking in dense clutter

Target tracking is a hot topic for unmanned aerial vehicle surveillance. Recently, the novel random sample consensus (RANSAC) algorithm shows a good tracking performance in dense clutter environment. However, the heavy computational burden limits the usage for unmanned aerial vehicle (UAV). In this paper, a density-based recursive random sample consensus (DBR-RANSAC) algorithm is proposed, which utilizes the density property of measurements within several steps to direct sampling. In the DBR-RANSAC, the randomness of sampling can be avoided and the computation complexity can be reduced particularly in dense clutter. The simulation results show the validity of the proposed algorithm.