摘要
提出一种改进后的应用于智能建筑门禁系统的人脸识别与报警的方法。阐述了人脸图像的预处理、数据压缩、分类器设计、安防报警的原理及实现过程。识别结果可与安防报警系统联动。实验数据表明,该方法的人脸识别效果和报警正确率有所提高。
An improved method of face recognition and alarm applied in access control system in intelligent building was presented. The theory and realization process of the pre-process of face image, data compression, design of classifier and safety alarm were discussed. The recognition result could be applied in safety alarm linkage system. The result of experiment showed that the method raised the accurate rate of face recognition effects and the accurate proportion of safety alarm.
出处
《低压电器》
北大核心
2009年第2期22-26,34,共6页
Low Voltage Apparatus
基金
建设部科研基金项目(2007-K03-04)
关键词
智能建筑
安防报警系统
数字监控系统
奇异值分解压缩
贝叶斯决策
RBF神经网络
人脸识别
intelligent building safety alarm system
digital monitoring system
singular value decomposition compression
Bayesian decision
RBF neutral network
face recognition
