全部 标题 作者
关键词 摘要

OALib Journal期刊
ISSN: 2333-9721
费用:99美元

查看量下载量

相关文章

更多...

基于改进CSSD的脑电信号特征提取方法

Keywords: 运动想象脑电,特征提取,共空域子空间分解,正则化方法,小样本

Full-Text   Cite this paper   Add to My Lib

Abstract:

针对脑-机接口系统在训练样本较少的情况下,存在脑电(EEG)信号特征值稳定性低、特征向量区分度差等不足,提出一种脑电特征提取方法,即正则化共空域子空间分解法(R-CSSD).该方法在传统共空域子空间分解(CSSD)算法的基础上引入正则化思想,通过正则化参数将目标实验者的训练数据与其他实验者(称为辅助实验者)的同类型训练数据进行有效结合,以构造正则化空间滤波器,完成对目标实验者运动想象EEG信号的特征提取,并进一步选用K近邻(KNN)算法实现脑电数据的分类.实验结果表明:在小训练样本情况下,R-CSSD方法有效提高了脑电信号特征值的稳定性,在提高分类正确率、降低时间消耗方面具有良好的性能.

References

[1]  肖庭延,于慎根,王彦飞.反问题的数值解法[M].3版.北京:科学出版社,2010:23-24.
[2]  杨立才,李佰敏,李光林.脑-机接口技术综述[J].电子学报,2005,33(7):1234-1241.YANG Li-cai,LI Bai-min,LI Guang-lin.A review ofbrain-computer interface technology[J].Acta ElectronicaSinica,2005,33(7):1234-1241.(in Chinese)
[3]  高上凯.浅谈脑-机接口的发展现状与挑战[J].中国生物医学工程学报,2007,6(26):801-803.GAO Shang-kai.Comments on recent progress andchallenges in the study of brain-computer interface[J].Chinese Journal of Biomedical Engineering,2007,6(26):801-803.(in Chinese)
[4]  李明爱,王蕊,郝冬梅,等.想象左右手运动的脑电特征提取及分类研究[J].中国生物医学工程学报,2009,28(2):166-170.LI Ming-ai,WANG Rui,HAO Dong-mei,et al.Featureextraction and classification of mental EEG for motorimagery[J].Chinese Journal of Biomedical Engineering,2009,28(2):166-170.(in Chinese)
[5]  尧德中,刘铁军.基于脑电的脑-机接口关键技术和应用前景[J].电子科技大学学报,2005,33(7):550-554.YAO De-zhong,LIU Tie-jun.Electroencephalogram basedbrain-computer interface:key techniques and applicationprospect[J].Journal of University of Electronic Scienceand Technology of China,2005,33(7):550-554.(inChinese)
[6]  LEDOIT O,WOLF M.A well-conditioned estimator forlargedimensional covariance matrices[J].J MultivariateAnal,2004,88(2):365-411.
[7]  李同磊,刘伯强,李可.基于脑电信号的手指动作识别[J].山东科学,2006,19(1):1-5.LI Tong-lei,LIU Bo-qiang,LI Ke.Finger movementbased on EEG recognition[J].Shandong Science,2006,19(1):1-5.(in Chinese)
[8]  刘琳,魏庆国.CSSD+AAR模型在脑电信号处理中的应用[J].通信技术,2009,10(42):207-210.LIU Lin,WEI Qing-guo.Application of CSSD and AAR inEEG signal processing[J].Communications Technology,2009,10(42):207-210.(in Chinese)
[9]  WANG Yi-jun,GAO Xiao-rong,GAO Shang-kai,et al.BCI competition 2003—data setⅣ:an algorithm based onCSSD and FDA for classifying single-trial EEG[J].IEEETrans Biomed Eng,2004,51(5):1081-1086.
[10]  TOMIOKA R,AIHARA K.Classifying matrices with aspectral regularization[C]∥The 24th Annual InternationalConference.Oregon:Oregon State University Corvallis,2007:895-902.
[11]  MORITZ G.Beamforming in noninvasive brain-computerinterfaces[J].IEEE Trans Biomed,2009,4(56):1081-1086.
[12]  李颖杰,邱意弘,朱贻盛.脑电信号分析方法及其应用[M].北京:科学出版社,2009:3-4.
[13]  FRUEDMAN J H.Regularized discriminant analysis[J].J Amer States Assoc,1989,84(405):165-175.
[14]  LU H.Boosting LDA with regularization on Univ.ofToronto,MPCA features for gait recognition[C]∥Biometrics Symp.Toronto:Univ.of Toronto,2007:1-6.
[15]  谢美华,王正明.用偏微分方程作图像分析与处理[J].激光与光电子学进展,2005,42(8):36-40.XIE Mei-hua,WANG Zheng-ming.Application of partialdifference equation in image analysis and processing[J].Laser and Optoelectronics Progress,2005,42(8):36-40.(in Chinese)
[16]  徐大宏.基于正则化方法的图像复原算法研究[D].北京:国防科学技术大学电子科学与工程学院,2009.XU Da-hong.Research of image restoration algorithmbased on regularization[D].Beijing:School of ElectronicScience and Engineering,National University of DefenseTechnology,2009.(in Chinese)
[17]  张永平,郑南宁,赵荣椿.基于变分的图像恢复算法及收敛性[J].自动化学报,2002,28(5):673-679.ZHANG Yong-ping,ZHENG Nan-ning,ZHAO Rong-chun.Based on variation and its convergence algorithmfor image restoration[J].Acta Automatica Sinica,2002,28(5):673-679.(in Chinese)
[18]  张胜,王蔚.基于CSSD和SVM的抑郁症脑电信号分类[J].中国生物医学工程学报,2008,6(27):827-831.ZHANG Sheng,WANG Wei.Application of CSSD andSVM for EEG signal classification[J].Chinese Journal ofBiomedical Engineering,2008,6(27):827-831.(inChinese)
[19]  刘笑嶂.人脸识别中线性判别分析的单参数正则化方法[J].计算机仿真,2008,25(10):215-218.LIU Xiao-zhang.A single-parameter-regularized methodin linear discriminant analysis for face recognition[J].Computer Simulation,2008,25(10):215-218.(inChinese)
[20]  叶柠,孙宇舸,王旭.基于共空间模式和K近邻分类器的脑-机接口信号分类方法[J].东北大学学报:自然科学版,2009,8(30):1107-1110.YE Ning,SUN Yu-ge,WANG Xu.Classification ofbrain-computer interface signals based on common spatialpatterns and K-nearest neighbors[J].Journal ofNortheastern University:Natural Science,2009,8(30):1107-1110.(in Chinese)

Full-Text

Contact Us

service@oalib.com

QQ:3279437679

WhatsApp +8615387084133