序列图像背景重建方法

研究了基于Kalman滤波理论的有色噪声滤波,探讨了其在序列图像背景重建中的应用。首先针对光线变化对背景的影响,建立背景的二阶滤波方程和观测方程,然后给出有色噪声滤波模型,利用Kalman滤波器对背景进行预测更新。试验结果表明,该方法能迅速更新背景,对光线变化具有良好的适应能力。

Background Rebuilding Method from Image Sequence

The problem of the colored noise filtering based on Kalman filter theory is studied and a method for rebuilding background image is discussed.Considering the effect of the change of illumina- tion condition on background,the two-order filter equation and the measurement equation of the back- ground are established,then a kind of Kalman fihering model for the colored noise is given,Kalman fil- ter is used to predict and update background.Experiments show that the method can update the back- ground quickly,thus having good adaptability to the light change.