智能监控系统中基于质心特征模型的动态场景检测

Dynamic Scene Detection Based on Centroid Characteristic Model for Intelligent Monitor System

在传统图像预处理与运动检测方法的基础上,提出了将数学形态学的膨胀腐蚀算子用于运动物体的连接填充,并在此基础上建立质心特征模型求运动物体的累积变化图像,实现运动物体检测,实验结果分析表明了该方法的有效性。

Extended corrosive arithmetic operators derived from mathenmtical morphology are brought forward on the connecting and filling of mobile image, which are based on traditional image pre manipulation and motion detection technique. In order to realize the mobile detection, the centroid characteristic models are created according to this theory to get the cumulated transformation image of mobile. The theory is approved by the experiment.