An Efficient Foreground Detection Algorithm for Visual Surveillance System

Date Added: Jun 2009
Format: PDF

This paper proposes an efficient foreground detection algorithm based on a new color space model and morphological filtering. This model uses each pixel's color distortion and brightness distortion to detect the candidate foreground pixels. The color distortion considers the vector's position in color space so that it can assemble the color features effectively. Also shadow elimination process removes the moving shadow. By applying it to background subtraction, it gets a comparative complete foreground object. For complex background, the post-processing and the filtering stage helps achieve better results. It has tested the presented algorithm with a variety of videos with various background and results clearly show the efficiency of the algorithm proposed.