Object Tracking in Video Pictures

OData support
Lois László Dr.
Department of Networked Systems and Services

The goal of this paper is to give an overview of object tracking and the problems connected closely to it and to present the development of an object tracking application. Connected problems like finding the object and tracking with multiple cameras, and some of their solutions are introduced. A classification of trackers with some examples are presented. The comparsion of the trackers’ most important abilities is shown. The operation of the tracker in a few cases is also presented. The goal of the application developed is to localize the objects having a user-defined color on the image at the beginning and track them on the following video frames. The objects are localized by a method develeoped by myself. The function assumes that the user has chosen the number of objects to be found on the image. Its operation’s main characteristic is that it chooses pixels having a H value in range of a predefined interval in the HSV color space. The tracker is based on the CAMSHIFT algorithm with use of Kalman filter (both found in the literature). This program is to be undergone more development in the future. It’s purpouse is to help people with learning medieval fencing.


Please sign in to download the files of this thesis.