There are many techniques for modeling and characterizing system behaviors. These techniques always require a sequence of steps such as acquisition of training data, training of the model and verification of the trained model performance. In this paper, we present a Matlab/Simulink-based simulator for a video-based target tracking system that is used to train AI learning models including artificial neural networks (ANN). It is shown that an ANN could emulate the eye-hand tracking (EHT) coordination of a human operator. The simulator-based experiments can be used to reveal many insights for training of the different ANN configurations for human EHT action. Its uses can be extended to system identification, adaptive control, machine learning and deep learning.