Robust approaches for the data association problem

We consider the problem of data association, which is central in multi-target tracking. Mathematically, this problem can be stated as a multidimensional assignment problem. Due to various sources of imprecision, the optimal solution of the multidimensional assignment problem can have low percentage of correct association. In order to improve the association correctness, we propose to evaluate the objective function coefficients by intervals instead of point values. This way, we obtain an interval-programming problem. Different strategies are proposed to tackle this new problem. Numerical examples for demonstrating the effectiveness of our approach are presented.