Computer Engineering and Applications ›› 2018, Vol. 54 ›› Issue (24): 149-155.DOI: 10.3778/j.issn.1002-8331.1709-0021

Previous Articles     Next Articles

EEG signals feature extraction based on EMD and CSP combined WOSF

ZHANG Xuejun1,2, WANG Longqiang1, HUANG Wanlu1, HUANG Liya1,2, CHENG Xiefeng1,2   

  1. 1.School of Electronic  and Optical Engineering, Nanjing University of Posts and Telecommunications, Nanjing 210023, China
    2.Nation-Local Joint Project Engineering Lab of RF Integration & Micropackage, Nanjing University of Posts and Telecommunications, Nanjing 210023, China
  • Online:2018-12-15 Published:2018-12-14



  1. 1.南京邮电大学 电子与光学工程学院,南京 210023
    2.南京邮电大学 射频集成与微组装技术国家地方联合工程实验室,南京 210023

Abstract: A feature extraction method based on Empirical Mode Decomposition(EMD), Common Spatial Pattern(CSP) and Wavelength Optimal Spatial Filter(WOSF) is proposed, Firstly, the EMD is used to decompose the EEG signal, and get a set of stationary time series called Intrinsic Mode Functions(IMFs). Secondly, selecting the appropriate IMFs for signal reconstruction, then the signal can be transformed into optimal signal through WOSF, the optimal signal is mapped to high-dimensional space through CSP, extracting the corresponding feature vector. Finally, the classification is performed using Support Vector Machine(SVM). After analyzing the result of the 9 subjects, the average accuracy classification rate obtained is over 95%, confirming the feasibility and availability of this method.

Key words: Empirical Mode Decomposition(EMD), Common Spatial Pattern(CSP), Wavelength Optimal Spatial Filter(WOSF), Intrinsic Mode Functions(IMF)

摘要: 基于经验模式分解和共空间模式,结合最优波长空间滤波,提出了三者相结合的特征提取方法。该方法首先利用经验模式分解进行分解,得到固有模态函数,选择合适的固有模态函数进行信号的重构,然后将重构的信号进行最优波长空间滤波变换,得到最优的波长选择信号,再经共空间模式投影映射,提取相应的特征向量,最后利用支持向量机进行分类。运用该方法对9位受试者进行分类结果分析,平均分类准确率在95%以上,实验表明,提出的算法具有较好的分类识别性。

关键词: 经验模式分解, 共空间滤波模式, 最优波长空间滤波, 固有模态函数