Čítání osob ve videosekvencích
Loading...
Downloads
9
Date issued
Authors
Adamec, Václav
Journal Title
Journal ISSN
Volume Title
Publisher
Vysoká škola báňská - Technická univerzita Ostrava
Location
Signature
Abstract
The aim of this thesis was to implement software for counting the persons in videosequences. The proposed algorithm uses an motion detection method Mixture of
Gaussians to segment the persons. People counting function is based on trajectory tracking using the Kalman filter. The template matching metod is used to increase the robust of application in the case of partial occlusion of observed individuals. The solution is implemented in C ++ programing language, using the OpenCV library.
Description
Import 22/07/2015
Subject(s)
counting the persons, trajectory tracking, Mixture of Gaussians, Kalman filter