基于多通道sEMG小波包分解特征的人手动作模式识别方法
|
摘要
为了满足主动康复训练和人机交互等复杂应用场景对多样性的人手运动模式识别需求,提出了一种基于多通道表面肌电信号sEMG小波包分解特征的人手动作模式识别方法;通过实验对比分析,确定了最佳采样布局方案,通过采集前臂表面肌电信号,设计了基于数字滤波器的肌电信号活动段自动标识算法,能快速准确地完成样本动作标签的制作;以原始肌电信号的小波包分解系数作为特征向量训练分类器;通过对比不同隐含层节点数对分类器模式识别准确率的影响,最终确定BP神经网络模式分类器的所有结构参数;设计并训练完成了BP神经网络人手运动模式分类器;对9种手部运动的平均识别率达到93.6%,计算时间小于150ms。
In order to meet the needs of complex application scenarios such as active rehabilitation training and human-computer interaction,a hand motion pattern recognition method based on multi-channel surface EMG signal(shorted as sEMG)wavelet packet decomposition is proposed.Through the comparison and analysis via experiments,the optimal sampling layout scheme is determined.The automatic identification algorithm of the active segment signal based on the digital filter is designed by collecting the sEMG signal of the forearm surface,and the production of the label can be completed quickly and accurately.The wavelet packet decomposition coefficient of the original sEMG signal is used as the feature vector training classifier.By comparing the influence of different hidden layer nodes on the accuracy of classifier pattern recognition,all the structural parameters of BP neural network model classifier are finally determined.The BP neural network for hand motion pattern classification was designed and trained.The average recognition rate of 9 kinds of hand movements was 93.6% and the calculation time was less than 150 ms.
In order to meet the needs of complex application scenarios such as active rehabilitation training and human-computer interaction,a hand motion pattern recognition method based on multi-channel surface EMG signal(shorted as sEMG)wavelet packet decomposition is proposed.Through the comparison and analysis via experiments,the optimal sampling layout scheme is determined.The automatic identification algorithm of the active segment signal based on the digital filter is designed by collecting the sEMG signal of the forearm surface,and the production of the label can be completed quickly and accurately.The wavelet packet decomposition coefficient of the original sEMG signal is used as the feature vector training classifier.By comparing the influence of different hidden layer nodes on the accuracy of classifier pattern recognition,all the structural parameters of BP neural network model classifier are finally determined.The BP neural network for hand motion pattern classification was designed and trained.The average recognition rate of 9 kinds of hand movements was 93.6% and the calculation time was less than 150 ms.
引文
[1]DE Luca C J,Adam A,Wotiz R,et al.Decomposition of surface EMG signals[J].Journal of neurophysiology,2006,96(3):1646-1657.
[2]Frigo C,Crenna P.Multichannel SEMG in clinical gait analysis:a review and state-of-the-art[J].Clinical Biomechanics,2009,24(3):236-245.
[3]Sugar T G,Ping H J,Kleneman E J,et al.Design and control of RUPERT:a device for robotic upper extremity repetitive therapy[J].IEEE Transactions on Neural Systems and Rehabilitation Engineering,2007,15(3):336-346.
[4]Perry J C,Powell J M,Rosen J.Isotropy of an upper limb Exoskeleton and the kinematics and dynamics of the human arm[J].Applied Bionics and Biomechanics,2009,6(2):175-191.
[5]Hargrove L J,Scheme E J,Englehart K B,et al.Multiple binary classifications via linear discriminant analysis for improved controllability of a powered prosthesis[J].IEEE Transactions on Neural Systems and Rehabilitation Engineering,2010,18(1):49-57.
[6]Ong SCW,Ranganath S.Automatic sign language analysis:A survey and the future beyond lexical meaning[J].IEEE Transaction on Pattern Analysis and Machine Intelligence,2005,27(6):873-891.
[7]Matsubara T,Morinoto J.Bilinear modelling of EMG signals to extract user independent features for multi-user myoelectric interface[J].IEEE Transactions on Biomedical Engineering,2013,60(8):2206-2213.
[8]李云.基于肌电模式的中国手语识别研究及康复应用探索[D].合肥:中国科学技术大学,2013.
[9]Smith A,Brown E E.Myoelectric control techniques for a rehabilitation robot[J].Applied Bionics and Biomechanics,2011,8(1):21-37.
[10]梅品高,罗志增.基于小波包分析和Elman网络的肌电信号处理[J].机电工程,2008,25:7-10.
[11]Al Omari F,Hui J,Mei C,et al.Pattern Recognition of eight hand motions using feature extraction of forearm EMG signal[J].Proceedings of the National Academy of Sciences,India Section A:Physical Sciences,2014,84(3):473-480.
[12]Futamata M,Nagata K,Magatani K.The evaluation of the discriminant ability of multiclass SVM in a study of hand motion recognition by using sEMG[A].Engineering in Medicine and Biology Society(EMBC),2012Annual International Conference of the IEEE[C].2012:5246-5249.
[13]Naik G R,Kumar D K,Jayadeve.Twin SVM for gesture classification using the surface electromyogram[J].IEEE Transactions on Information Technology in Biomedicine,2010,14(2):301-308.
[2]Frigo C,Crenna P.Multichannel SEMG in clinical gait analysis:a review and state-of-the-art[J].Clinical Biomechanics,2009,24(3):236-245.
[3]Sugar T G,Ping H J,Kleneman E J,et al.Design and control of RUPERT:a device for robotic upper extremity repetitive therapy[J].IEEE Transactions on Neural Systems and Rehabilitation Engineering,2007,15(3):336-346.
[4]Perry J C,Powell J M,Rosen J.Isotropy of an upper limb Exoskeleton and the kinematics and dynamics of the human arm[J].Applied Bionics and Biomechanics,2009,6(2):175-191.
[5]Hargrove L J,Scheme E J,Englehart K B,et al.Multiple binary classifications via linear discriminant analysis for improved controllability of a powered prosthesis[J].IEEE Transactions on Neural Systems and Rehabilitation Engineering,2010,18(1):49-57.
[6]Ong SCW,Ranganath S.Automatic sign language analysis:A survey and the future beyond lexical meaning[J].IEEE Transaction on Pattern Analysis and Machine Intelligence,2005,27(6):873-891.
[7]Matsubara T,Morinoto J.Bilinear modelling of EMG signals to extract user independent features for multi-user myoelectric interface[J].IEEE Transactions on Biomedical Engineering,2013,60(8):2206-2213.
[8]李云.基于肌电模式的中国手语识别研究及康复应用探索[D].合肥:中国科学技术大学,2013.
[9]Smith A,Brown E E.Myoelectric control techniques for a rehabilitation robot[J].Applied Bionics and Biomechanics,2011,8(1):21-37.
[10]梅品高,罗志增.基于小波包分析和Elman网络的肌电信号处理[J].机电工程,2008,25:7-10.
[11]Al Omari F,Hui J,Mei C,et al.Pattern Recognition of eight hand motions using feature extraction of forearm EMG signal[J].Proceedings of the National Academy of Sciences,India Section A:Physical Sciences,2014,84(3):473-480.
[12]Futamata M,Nagata K,Magatani K.The evaluation of the discriminant ability of multiclass SVM in a study of hand motion recognition by using sEMG[A].Engineering in Medicine and Biology Society(EMBC),2012Annual International Conference of the IEEE[C].2012:5246-5249.
[13]Naik G R,Kumar D K,Jayadeve.Twin SVM for gesture classification using the surface electromyogram[J].IEEE Transactions on Information Technology in Biomedicine,2010,14(2):301-308.