Abstract: Point to the problem that the automatic background updating can not be well realized when using traditional background subtraction method，a new fast adaptive background subtraction method is presented in this paper，which takes advantage of Minimum Mean Square Error（MMSE） theory to make estimation，then using corresponding algorithm to realize background updating under the cases that background is taking slowly changes or sudden changes.Testing results demonstrate that when the method is applied，the self-adaptive updating of background can be realized with a high accuracy，meanwhile it is simple to implement and suitable for real-time detection.

Key words: self-adaptive background updating, Minimum Mean Square Error（MMSE）, vehicle detection

摘要： 针对采用传统背景差法检测运动车辆时很难实现背景自动更新的问题，采用了快速自适应背景生成算法，并考虑到背景有缓变和突变两种情况，引入最小均方误差（MMSE）估计理论进行判断，针对不同的情况采取相应的算法实现背景更新。实验数据表明：采用该方法较以往准确性大大提高，同时达到了自适应背景更新的目的，并且算法实现简单，实时性较高。

关键词: 自适应背景更新, 最小均方误差, 车辆检测