A new insight to the ANN-based multitarget tracking data association

The method of joint probabilistic data association (JPDA) is most commonly used to solve the problem of tracking multiple targets in the presence of clutter because all the observations and all tracks are used to find the association probabilities, which are used as weighting coefficients in the formation of a weighted-average measurement for updating each track. The computation of association probabilities is quite complex, especially with the presence of clutter. An artificial neural network (ANN) method is a new technique for finding association probabilities as a contained optimization problem by minimizing the energy function which depends mainly on the five coefficients. A new insight into the affect of the five parameters on the miscorrelation of ANN-based data association is presented