Moving region detection based on background difference

Based on Gaussian mixture model and background difference, a novel moving object region detection method is presented for static camera surveillance. First of all, the adaptive threshold is used to get background difference image. Secondly, by analyzing the connected domain and setting a threshold, the influence of the noise in the difference image is eliminated and the moving object regions are determined. It can be used as moving window of the accurate detection for a subsequent object outline. The experimental results show that the algorithm can achieve moving object region detection of rigid or non-rigid in static camera surveillance.