Systém pro detekci osob v inteligentních budovách

Abstract

This diploma thesis deals with computer vision methods for detecting people from the video signal. Person Detection is designed for intelligent buildings, specifically indoor areas. The result of the work is a system designed for laboratory testing of person detection in intelligent buildings. In the first part of the thesis are theoretically analyzed selected known methods of person detecting in picture and subsequent image editing. In the next part, the selected methods are applied to the identical test videosequence of the images. In the final part of the thesis, the results of the detection of individuals of the applied methods are qualitatively evaluated by verification of the ideal detection result.