Title :
A model change detection approach to dynamic scene modeling
Author :
Kim, Seon Joo ; Doretto, Gianfranco ; Rittscher, Jens ; Tu, Peter ; Krahnstoever, Peter Tu Nils ; Pollefeys, Marc
Author_Institution :
Dept. of Comput. Sci., Univ. of North Carolina, Chapel Hill, NC, USA
Abstract :
In this work we propose a dynamic scene model to provide information about the presence of salient motion in the scene, and that could be used for focusing the attention of a pan/tilt/zoom camera, or for background modeling purposes. Rather than proposing a set of saliency detectors, we define what we mean by salient motion, and propose a precise model for it. Detecting salient motion becomes equivalent to detecting a model change. We derive optimal online procedures to solve this problem, which enable a very fast implementation. Promising results show that our model can effectively detect salient motion even in severely cluttered scenes, and while a camera is panning and tilting.
Keywords :
image motion analysis; object detection; background modeling; dynamic scene modeling; model change detection; pan-tilt-zoom camera; salient motion detection; salient scene motion; Cameras; Computer science; Computer vision; Focusing; Layout; Motion detection; Surveillance; Testing; Vehicle dynamics; Visualization; PTZ camera; background modeling; dynamic scene modeling; dynamic textures; focus-of-attention; linear dynamical systems; model change detection; sequential generalized likelihood ratio;
Conference_Titel :
Advanced Video and Signal Based Surveillance, 2009. AVSS '09. Sixth IEEE International Conference on
Conference_Location :
Genova
Print_ISBN :
978-1-4244-4755-8
Electronic_ISBN :
978-0-7695-3718-4
DOI :
10.1109/AVSS.2009.30