Efficient and accurate background subtraction using multilevel background model

Background subtraction is an essential step of intelligent video surveillance and hasgot a lot of interest among researcher community in recent years. It has a critical impacton the performance of object tracking and activity analysis. In this paper we proposea new multi-level background modeling to overcome dynamic background problem. Atrst, an image is segmented to foreground or background in larger window(12 12)level and results are rened by smaller windows(6 6 and3 3)level and pixel basedoperation. Our pixel based segmentation section uses VIBE method , which is fast andneeds less memory to conserve background model components, with some modicationto detect shadows. Subtraction is done in coarse levels rstly, and resulted foregroundare investigated more by smaller windows(ne levels). This makes the algorithm to bemore ecient. A new once-o background changing detection and model updating isproposed to make our algorithm as accurate as possible. The last part of our algorithmis enhancement where, we have used morphological operators in order to improve thesubtraction quality. The approach provides us with many advantages compared to thestate-of-the-art. Experimental results clearly justify our strategy