Research on Active Visual Tracking Based on Vestibulo-Ocular Reflex

Abstract

Tracking moving objects is an important research direction in the field of robot vision. Tracking failure is still a challenging problem when the object tracking system is disturbed by the outside world or itself. In this paper, based on the theory of vestibulo-ocular reflex and its adaptive characteristics, fuzzy control is introduced to compensate the control parameters of the system, which leads to reduction of the external interference to the system and improvement of the tracking control system of the object. By imitating and learning the mechanism of human visual system, the performance of object tracking can be improved. In this paper, a "head-eye" structure is built. By experimenting with different interferences on the system, the sight is fixed on the target, which verifies the effectiveness of the algorithm.