A new technique for background modeling and subtraction for motion detection in real-time videos

Abstract

Background modeling and subtraction is a core component in motion detection and analysis, using a stationary camera. This paper proposes a new technique for background modeling and subtraction capable of processing a real-time video and achieving high detection rates. Color information alone can't handle dynamic environments. On the other hand, edge or texture information alone is not sufficient for uniform regions. To overcome these limitations, the proposed technique integrates Local Binary Pattern (LBP) texture feature, RGB color feature and Sobel edge feature, using “Choquet” fuzzy integral to avoid the uncertainty in the classification. This is performed by extracting the color and the edge gray scale confidence maps and introducing the texture confidence map. Then, a single median filtering and a connected components algorithms are applied to remove the noise and label the connected regions. Experimental results for both indoor and outdoor dataset sequences confirmed the robustness and the effectiveness of the proposed technique against illumination variation and scene motion.