基于多传感器信息融合技术的智能视频监控系统研究  

Design of the smart video monitoring system based on multi-sensor information fusion

作　　者：刘鹏飞[1] 赵陈[1] 王嘉永[1] 张洁[2] 

机构地区：[1]兰州交通大学电子与信息工程学院,兰州730070 [2]河北师范大学汇华学院,石家庄050091

出　　处：《电子测试》2011年第4期38-40,共3页Electronic Test

摘　　要：介绍了一种基于多传感器信息融合技术的智能视频监控系统的信息融合设计。为了降低监控系统中的单一传感器的误报率和漏报率,通过对视频监控系统的人体红外、视频摄像等多种传感器的网络集成,将不同传感器采集的不同的描述信息进行了有效地特征提取和传输,并运用模糊神经网络的多传感器信息融合算法进行多次仿真实验。实现了该系统对检测区域的实时估计决策和报警,提高了智能视频监控系统的实时报警、智能识别及智能跟踪的准确性。This paper introduced a design of the smart video monitoring system based on multi-sensor information fusion.In order to lower the proportion of errors or omits which the single sensor generate,It is a neural network merger of video system and multi-sensors,and collects the characteristics of various sensors.In order to determine the surveyed area and alarm in real time,we do multi-run simulation according to using the multi-sensor information fusion of the blurred neural network.Thus has enriched the identification of smart video monitoring system,and added its tracking accuracy.

关 键 词：多传感器信息融合 视频监控 神经网络 

分 类 号：TP277[自动化与计算机技术—检测技术与自动化装置]