Distinguishing moving objects using Kalman Filter and Phase Correlation methods

A Neural network based on Kalman Filter and Phase Correlation is devised to recognize and distinguish objects moving in a plane. A bank of Kalman Filters in parallel are used to track the moving objects and the Phase correlation method is used to recognize the moving objects. Both methods together are used to distinguish identical objects based on Kalman estimates of the location and speed. Experiments were performed using MATLAB 2013a and it is seen errors occur when identical objects are occluded moving at similar speeds.