摘要
提出了一种基于机器学习的实时异常监测方案,VFW技术的使用减少了对视频设备的依赖;具备机器学习的功能,能有效区别背景噪声和入侵者;采用了基于背景减法的运动检测算法,能在廉价的家用PC机上通过廉价的摄像头实现实时监视报警并保存现场信息．
A real time scheme of abnormity detection based on machine study was brought forward and the use of VFW reduced the reliance on video equipment.The system can adapt to environment automaticly to discriminate between background noise and intruder. By using the simple arithmetic of motion detection based on the background subtration, the system can run on a cheap PC, either giving an alarm or saving the context when something unexpected intrudes.
出处
《包头钢铁学院学报》
2005年第4期325-329,共5页
Journal of Baotou University of Iron and Steel Technology
关键词
VFW视频捕获
机器学习
背景减法
异常监测
VFW video capture
machine study
background subtration
abnormity detection
