普适环境下基于脑电的身份及上下文状态识别系统的设计与研究.pdfVIP

普适环境下基于脑电的身份及上下文状态识别系 统的设计与研究 刘泉影1 毛承胜1 聂碧娟2 胡斌1,3 (1 兰州大学 计算机软件与理论系，兰州，730000 ) (2 武汉工程职业技术学院 经贸系，武汉，430080) (3 英国伯明翰城市大学 普适感知计算中心，伯明翰，B42 2SU) 摘要: 为了研究基于脑电的身份识别的具体实现，并满足用户对于普适和上下文状态感知的需求， 本文提出了一种基于脑电的身份及上下文状态识别系统。本文针对一局域范围内身份识别的普适 化需求（如各接口身份验证），提出一种基于脑电的身份及上下文状态识别方法的普适化实现方 案。并且，通过实验采集到5 个受试者3 种上下文状态的脑电信号，经过脑电去噪、脑电特征提 取和kNN(k=1)分类之后，分别得到96.43%的身份识别正确率和92.31%的上下文状态识别正确率， 由此证实了此方案的可行性。 关键词:普适；身份识别；上下文状态识别；EEG 中图分类号: U459. 2 文献标识码: A The Design and Research on EEG-based individual and context recognition system in ubiquitous environment 1 1 2 3 Quanying Liu , Chengsheng Mao , Bijuan Nie , Bin Hu1 (1 Department of Computer Science and Technology, Lanzhou University, Lanzhou 730000, China) (2 Economy and Trade, Wuhan Engineering Institute, Wuhan 430080, China) (3 Intelligent Contextual Computing Group in Pervasive Computing Centre, the Birmingham City University, Birmingham B42 2SU, United Kingdom) Abstract: Because of the irreplication and individual specificity of the electroencephalogram (EEG), EEG based individual recognition is hitting in the biometrics. With the improvement of wireless sensors, ubiquitous computing is becoming practice. However, most of the existing individual identification system cannot fulfill the ubiquitous requirement, nor aware context. In this paper, aiming at meeting the ubiquitous requirement, as well as context recognition, in a certain region, we present a EEG-based individual and context recognition ubiquitous solution. In addition, ex