Sledování zdravotního stavu řidiče pomocí obrazů

Abstract

This bachelor thesis is about tracking driver’s health care. Implemented software is focused on monitoring and analyzing data of face area. Algorithm is designed so it obtains driver’s facial landmarks and its used to get driver’s health condition. Algorithm uses HOG and shape predictor trained on dataset 300-W. For one of the symptoms is also used Support Vector Machine. To test functinality and succes rate I created application i C++ language using libraries such as OpenCV and Dlib.