一种改进的视频监控背景更新算法

An Improved Algorithm for Background Updating in Video Surveillance

在智能视频监控中,背景及时准确地更新是监控系统自动检测运动目标的关键。针对ⅡR滤波更新算法在实际应用中的缺陷,提出了基于当前帧、背景帧与初始背景帧三帧信息相结合的方法更新背景。实验表明,该方法对光变具有良好的鲁棒性,噪声小,速度快,具有较高的实用价值。

A timely and accurate background updating algorithm is a key point for the surveillance system to detect the moving objects automatically in intelligent video surveillance.In viewing the defect of the iterative weighted background updating algorism in practice,a new algorithm based on the current frame and its corresponding background frame and the initial background frame are proposed.Experiments show that this algorithm is robust under lighting variations,noiseless,fast in speed and has great value in practice.