Detekce očního bělma v obrazech

Abstract

This thesis deals with problematics of sclera detection inside images. The main aim of this work is describe of available sclera detection methods, realize one useful algorithm and verify it on on accessible samples. For implementation of application is utilized programming language Python with OpenCV library and Haar Cascade Classifiers, which are necessary for eyes and face detection. In thesis are described methods on thresholding bases and contour detection, algorithm for finding similar colored pixels and facial landmarks predictor. Part of this thesis is also comparison of success of individual detection