Research on Moving Target Detection Based on ViBe Algorithm

Abstract

ViBe algorithm is an efficient pixel level video background modeling algorithm. Aiming at the problem of ghost and dynamic background false detection in traditional ViBe algorithm, an improved algorithm based on improved ViBe algorithm is proposed. Firstly, the initial background model and the background update algorithm of three frame difference are constructed based on the time information to solve the problem of ghost elimination in the ViBe algorithm. Aiming at the problem of false detection in the process of dynamic background detection in the ViBe algorithm, an adaptive threshold dynamic background motion detection algorithm combined with the three frame difference method is proposed, Three frame difference method is used to determine the dynamic interference area, and adaptive threshold motion detection is used to detect the dynamic interference area. The algorithm can quickly remove ghosts and reduce the probability of false detection in dynamic background. Experimental results show that the algorithm is effective and robust.