A Discrimination Method of Ghost and Static Target Based on Improved VIBE
Author:Shengnan Guo,Chunxue Wu,Xiao Lin,Yan Wu,Jun Ai Author Unit: School of Optical-Electrical & Computer Engineering, University of Shanghai for Science & Technology, Shanghai 200093, China School of public and environmental affairs, Indiana University Bloomington, 47405, USA
Abstract:Accurate foreground detection is the key to ensuring embedded real-time monitoring systems. In order to
quickly eliminate ghosts and detect stationary targets more accurately, this paper proposes a novel foreground
detecting method based on improved Vibe. Firstly, the foreground is detected by the original Vibe algorithm,
and stationary foreground region and moving object are distinguished by joining foreground Time out map
(TOM) mechanism. Then, it judges whether a region is a ghost region or a stationary target region based on
the variance of the pixel values of the stationary foreground region. Finally, different strategies are adopted to
eliminate ghost and retain static object. A series of experiments illustrate that the proposed method is superior
to original Vibe and GMM in eliminating ghost and retaining static object.
Keywords:Discrimination method; Ghost and static target based; VIBE
Download information [SIZE:1.80 MB Amount of downloads: second]