视频监控系统中一种运动目标的检测方法  

Motion Detection Algorithm in Video Surveillance System

作　　者：孙涛[1] 陈瑞平[1] 

机构地区：[1]烟台大学,山东烟台264005

出　　处：《电视技术》2012年第7期127-129,148,共4页Video Engineering

摘　　要：提出了一种基于改进的背景差分法的运动目标检测和识别的方法,该算法用于视频监控系统中运动目标检测和报警。双阈值法和动态阈值法有效地检测出图像中的运动目标。Matlab 7.0中对算法进行了仿真,实验表明,该方法有效去除了运动目标阴影及背景噪声,可准确地检测出运动目标。A method of moving object detection and recognition is proposed based on the background subtraction algorithm. It is used in an automatic moving object detection algorithm applied to real-time video surveillance system. The double thresholds can detect intruders with high quality and stead- y. Simulation is done for this model in Matlab 7. 0. The experiment shows that using the method, the shadow of moving object and noises can be removed from background effectually, and moving object can be detected exactly.

关 键 词：视频监控 背景差分法 运动目标检测 迭代法 

分 类 号：TN919.8[电子电信—通信与信息系统] TP391.41[电子电信—信息与通信工程]