master thesis

SignatureMA 016
AuthorRafael Rodríguez
TitleBackground Subtraction with PTZ Cameras
TutorDipl-Ing. Rubén Heras Evangelio
ProfessorDr.-Ing. Thomas Sikora
DateMarch 2012
AbstractThe segmentation of moving objects is often used as a first step in many video applications such as traffic monitoring and human activity understanding. Background subtraction is a commonly adopted technique for this segmentation, which basically involves detecting the moving objects by the difference between the current frame and a background model. However, most of the background subtraction approaches proposed in the literature are restricted to scenarios with the static cameras due to the need of a background model, being therefore not applicable in scenarios with Pan-Tilt-Zoom (PTZ) cameras, which are popular option because of their ability to cover a wide field of view.
Key wordsbackground subtraction, image registration, interpolation methods, mosaic