Moving Object Detection Method Based on Spatio-Temporal Background Model

Abstract

The existing background modelling methods mostly only use the temporal or spatial information of pixels, so that the detection precision is decreased. To solve this problem, a background modelling method fusing spatio-temporal information of pixels is proposed in this paper. Firstly, the proposed method initialized the temporal background model using the historical observations of pixels. Meanwhile, the method established the spatial background model by means of sampling pixel values from the 8-connected neighborhood of pixels. And then in the process of moving object detection, the temporal model was updated in the way of FIFO (first in first out), and the spatial model was updated using the random strategy. The experimental results indicate that moving objects can be detected effectively by the proposed method in real time. Specifically, it can weaken the effects caused by dynamic background and illumination changes. Besides, it can inhibit the interference of camera jitter to some extent.