Adaptive probabilistic data-association algorithm for tracking in cluttered environment

摘要

A probabilistic data-association filter which uses adaptive update times, for tracking a target in a cluttered environment and in an environment where the probability of detection (P/sub d/) is less than the unity, is presented and compared with the probabilistic data-association filter which uses a constant update time. The tracking performance of the algorithm is assessed by Monte Carlo simulation on different target trajectories. A comparison is made using the percentage of tracks lost, the mean-square errors, the mean update times and mean number of updates.