Search

Talcing account of the problems that the standard probability hypothesis density filter is difficult to estimate the target state correctly and has high computational complexity when there are multiple targets move close to each other in clutter environments, this paper put forward a fast close-spaced target tracking algorithm based on the probability hypothesis density. The proposed algorithm firstly used an adaptive gate technology to divide the measurement set originated from the real targets from the sensor measurement set. And then used the real target measurement set to update the predicted intensity. Finally, a detection-guided close-spaced target reweight approach selectively redistributed the inaccurate component weights within the posterior intensity at each discrete time. Experimental results illustrate that the proposed algorithm not only has target tracking with high efficiency, but also has good stability. [ABSTRACT FROM AUTHOR]