运动检测中的背景建立和更新

Background Updating and Setting in the Moving Detection

在传统的差分算法基础上,提出了一种新的背景的建立和更新的解决方法。在背景建立阶段提出了在空域基于象素灰度出现频率最大值的改进快速算法,完成原始背景的建立;在检测阶段,采用自适应Kalman滤波器参数学习完成当前背景的更新。

Based on the traditional difference algorithm, a new background construction and updating method is proposed. During the background construction phase, a quick algorithm based on the max gray frequency is proposed. During the detecting phase, adaptive Kalman-Filter learning algorithm is applied to update the current background.