A modification of a proposed model for the eye tracking movements

A sampled-data model for the eye tracking movements has been recently suggested by L. Young (1962, 1963). It was stated that for step inputs only the part of the model which corresponds to the saccadic system will respond and the part corresponding to the pursuit system will remain inactive due to a velocity limiter of about 30° per second. However, a closer consideration of the model reveals that this is not the case. The velocity estimator samples at intervals of 0.2 seconds, hence, for a step input less than 0.2 X 30 = 6° the pursuit system wiii try to follow. This is not in agreement with the physiological experimental result that even for step inputs as low as 1° the saccadic system is the only one operating. Also, the velocity limiter must be at least at 25° per second. Hence, a change of the numerical values of the parameters of the model cannot improve it. This communication does not intend to present a complete model for the eye tracking movements but simply to suggest some improvements on the one proposed by Young. However, one can already see that the more suitable approach for a complete model would be by using neural nets rather than lumped elements.