A novel motion detection approach based on the improved ViBe algorithm

Ghost elimination is considered as a vital phase towards the moving object detection problem. Despite all the efforts of the existing methods have been made so far, finding an accurate and fast computational approach to solve ghost elimination is still a challenging problem for moving detection. In this paper, we propose a novel motion detection approach based on the ViBe algorithm. We employ the weighted window to select the dynamic adaptive optimal parameter for the ViBe model firstly. After that, the model integrates the inter-frame difference algorithm to accelerate the process of the ghost elimination. Experiment results show that our approach can detect the moving objects in real-time and get the higher accuracy result in the comparisons with the existing approaches.