An ANN approach to multisensor multitarget tracking

In this paper, the multisensor multitarget tracking problem is considered. The data association is most important and difficult in the multitarget tracking environment. We first formulate it with the maximum likelihood method, and then we use a random stochastic ANN and modified mean field annealing to solve the optimal solution. Compared with other nonlinear optimization algorithms, this method can achieve the global optimal solution and it needs fewer iteration times.<
>