Focus of expansion estimation with a neural network

This paper presents an approach using a Hopfield neural network to solve the focus of expansion location problem given a set of 2D input motion vectors. The problem is formulated as the minimization of the energy function of a 2D Hopfield neural network, whose minimum value corresponds to the best solution of the problem. Results on both synthetic and real 2D motion maps shows that the method is robust and tolerant to noise and small rotational components in the input data