Tracking problem for eye movement

Abstract

The eyes move in order to place an intended object on the region of the retina with greatest visual acuity, called the fovea. Movement of eyes are produced via the coordinated action of several muscle groups acting in response- to neurological signals. The focus of this thesis is to study a particular type of eye movement called smooth pursuit from the perspective of control theory. There are several types of eye movements reported in the research literature. Primarily among them are the (1) saccadic or fast eye movement subs\stem (2) pursuit or tracking subsystem (3) vergence subsystem (4) vestibular subsystem. In contrast to the saccadic eye movement research, scant attention has been paid to the study of the remaining types from the viewpoint of control. Here a detailed description is given on how to design a controller for the pursuit eye movement using an ocular motor plant model from recent research literature. This model represent capabilities of muscle actuators of the eyes reasonably accurately from the viewpoint of their actuator and length limitations and other nonlinearities, but disregards details of cortical circuitry behind signal generation. Performance of the designed controller is demonstrated via simulations pertinent to pursuit of sinusoidal, ramp and parabolic signals. It is argued that the designed controller performs better than the actual occular motor plant for the reason that the additional restrictions imposed by cortical circuitry are completely disregarded in the design. Overall, the- latency and peak velocity of the designed controller is shown to be within a factor of four of the actual occular motor controller during pursuit.