Irregular patterns learning and matching in an example vision system

A method of irregular patterns learning and matching in an example computer vision system designed for eye tracking is presented. The main considered issue is the problem of patterns learning. In the presented solution it requires creation of a set of patterns with some kind of dynamic management. An adaptation of the set of patterns was achieved in that way that a new pattern may be added to the set of patterns and next undergoes an evolution during the process of matching. The matching method is based on the Hough transform for irregular patterns. In order to guarantee the performance the process of patterns matching is supported by a histogram analysis.