This paper reviews and evaluates performance of few common background subtraction algorithms which are medianbased, Gaussian-based and Kernel density-based approaches. These algorithms are tested using four sets of image sequences contributed by Wallflower datasets. They are the image sequences of different challenging environments that may reflect the real scenario in video surveillances. The performances of these approaches are evaluated in terms of processing speed, memory usage as well as object segmentation accuracy. The results demonstrate that Gaussian-based approach is the best approach for real-time applications, compromising between accuracy and computational time. Besides, this paper may provide a better understanding of algorithm behaviours implemented in different situation for real-time video surveillance applications.

This article presents a new method for background subtraction (BGS) and object detection for a real-time video application using a combination of frame differencing and a scale-invariant feature detector. This method takes the benefits of background modelling and the invariant feature detector to improve the accuracy in various environments. The proposed method consists of three main modules, namely, modelling, matching and subtraction modules. The comparison study of the proposed method with a popular Gaussian mixture model proved that the improvement in correct classification can be increased up to 98% with a reduction of false negative and true positive rates. Beside that the proposed method has shown great potential to overcome the drawback of the traditional BGS in handling challenges like shadow effect and lighting fluctuation.

No abstract