详细信息
基于能量特征的左右手运动想象脑信号的识别方法 ( EI收录)
Recognition of Right and Left Motor Imagery Based on Energy Features
文献类型:期刊文献
中文题名:基于能量特征的左右手运动想象脑信号的识别方法
英文题名:Recognition of Right and Left Motor Imagery Based on Energy Features
作者:金晶[1];王行愚[1];张秀[1]
机构:[1]华东理工大学信息科学与工程学院,上海200237
年份:2007
卷号:33
期号:4
起止页码:536
中文期刊名:华东理工大学学报(自然科学版)
外文期刊名:Journal of East China University of Science and Technology
收录:CSTPCD;;EI(收录号:20073810818853);Scopus;北大核心:【北大核心2004】;CSCD:【CSCD2011_2012】;
基金:国家自然科学基金项目(60543005;60674089)
语种:中文
中文关键词:小波;支持向量机;自适应自回归系数;相同步分析法
外文关键词:wavelet; SVM; ad slty apti of Technology in 2003. ve autoregressive parameters; phase synchronization analysis
摘要:提出了一种基于能量特征的左右手运动想象识别方法,利用快速傅里叶变换分析特定脑电(μ波和β波)的频率分布,然后利用小波分解去噪,再利用小波包分析脑电能量,提取能量特征,最后基于支持向量机(SVM)进行左右手运动想象的识别。本文把能量作为特征的支持向量机(SVM)识别法分别与自适应自回归系数法(AAR)和相同步分析法进行比较。仿真结果表明:在相同样本数据情况下,能量特征作为特征向量的SVM识别准确率明显高于其他2种方法。
We present a method to recognize the electroencephalogram (EEG) signal of right and left motor imagery based on features of energy. We utilize fast Fourier transform to analyze the frequency range of μ and β. We make use of wavelet to denoise and wavelet packets to analyze the energy of EEG and get its features. Then we recognize the EEG signal of right and left motor imagery based on SVM. The results of simulation indicate that the SVM method based on energy features is better than the method of adaptive autoregressive parameters and the method of phase synchronization analysis when we take the same data provided by Graz University of Technology in 2003.
参考文献:
正在载入数据...
