Many computer vision algorithms such as object tracking and event detection assume that a background model of the scene under analysis is known. However, in many practical circums...
A robust video object segmentation algorithm for complex conditions in surveillance systems is proposed in this paper. This algorithm contains an unsupervised K-Means background c...
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...
Moving object detection is very important for video surveillance. In many environments, motion maybe either interesting (salient) motion (e.g., a person) or uninteresting motion (...
Abstract. Background updating is an important aspect of dynamic scene analysis. Two critical problems: sudden camera perturbation and the sleeping person problem, which arise frequ...