Detecting regions of interest in video sequences is the most important task in many high level video processing applications. In this paper a robust technique based on recursive l...
Abstract. Robust foreground object segmentation via background modelling is a difficult problem in cluttered environments, where obtaining a clear view of the background to model i...
Vikas Reddy, Conrad Sanderson, Andres Sanin, Brian...
Background subtraction is commonly used to detect foreground objects in video surveillance. Traditional background subtraction methods are usually based on the assumption that the...
We present a new method to robustly and efficiently analyze foreground when we detect background for a fixed camera view by using mixture of Gaussians models and multiple cues. Th...
In this paper, we propose a panorama-based object tracking scheme for wide-view surveillance systems that can detect and track moving objects with a pan-tilt camera. A dynamic mosa...