Detekce míry otevření očí a frekvence mrkání ve videosekvencích

Abstract

This master thesis is focused on designing the methods for detection of the measure of eye opening and frequency of blinking. Detection of iris, eyelids and classification using SVM are used to detect eye states. By combining these detections, eye blinking detection is improved. The measure of eye opening is determined from detected distance of eyelids. Proposed methods were experimentally evaluated in four videosequences of three different subjects in varying lighting conditions. Implementation is done in C++.