Multi-feature gait recognition with DNN based on sEMG signals

Institute of Intelligent Control and Robotics, School of Automation, Hangzhou Dianzi University, Hangzhou 310018, China

Academic Editor: Mehdi Sookhak

Received: 02 February 2021 Accepted: 20 April 2021 Published: 23 April 2021

This study proposed a gait recognition method based on the deep neural network of surface electromyography (sEMG) signals to improve the stability and accuracy of gait recognition using sEMG signals of the lower limbs. First, we determined the parameters of time domain features, including the mean of absolute value, root mean square, waveform length, the number of zero-crossing points of the sEMG signals after noise elimination, and the frequency domain features, including mean power frequency and median frequency. Second, the time domain feature and frequency domain feature were combined into a multi-feature combination. Then, the classifier was trained and used for gait recognition. Finally, in terms of the recognition rate, the classifier was compared with the support vector machine (SVM) and extreme learning machine (ELM). The results showed the method of deep neural network (DNN) had a better recognition rate than that of SVM and ELM. The experimental results of the participants indicated that the average recognition rate obtained with the method of DNN exceeded 95%. On the other hand, from the statistical results of standard deviation, the difference between subjects ranged from 0.46 to 0.94%, which also proved the robustness and stability of the proposed method.
- gait recognition,
- multi-feature fusion,
- deep neural network (DNN),
- robustness and stability
Citation: Ting Yao, Farong Gao, Qizhong Zhang, Yuliang Ma. Multi-feature gait recognition with DNN based on sEMG signals[J]. Mathematical Biosciences and Engineering, 2021, 18(4): 3521-3542. doi: 10.3934/mbe.2021177

Related Papers:

Abstract

This study proposed a gait recognition method based on the deep neural network of surface electromyography (sEMG) signals to improve the stability and accuracy of gait recognition using sEMG signals of the lower limbs. First, we determined the parameters of time domain features, including the mean of absolute value, root mean square, waveform length, the number of zero-crossing points of the sEMG signals after noise elimination, and the frequency domain features, including mean power frequency and median frequency. Second, the time domain feature and frequency domain feature were combined into a multi-feature combination. Then, the classifier was trained and used for gait recognition. Finally, in terms of the recognition rate, the classifier was compared with the support vector machine (SVM) and extreme learning machine (ELM). The results showed the method of deep neural network (DNN) had a better recognition rate than that of SVM and ELM. The experimental results of the participants indicated that the average recognition rate obtained with the method of DNN exceeded 95%. On the other hand, from the statistical results of standard deviation, the difference between subjects ranged from 0.46 to 0.94%, which also proved the robustness and stability of the proposed method.

References

[1]	M. Perc, The dynamics of human gait, Eur. J. Phy., 26 (2012), 525-534.
[2]	C. Mei, F. R. Gao, Y. Li, A determination method for gait event based on acceleration sensors, Sensors, 19 (2019), 5499.
[3]	J. Ryu, B. Lee, D. H. Kim, sEMG signal-based lower limb human motion detection using a top and slope feature extraction algorithm, IEEE Signal Process. Lett., 24 (2017), 929-932. doi: 10.1109/LSP.2016.2636320
[4]	P. Connor, A. Ross, Biometric recognition by gait: a survey of modalities and features, Comput. Vis. Image Underst., 167 (2018), 1-27. doi: 10.1016/j.cviu.2018.01.007
[5]	J. S. Richman, M. J. Randall, Physiological time-series analysis using approximate entropy and sample entropy, Am. J. Physiol-Heart C., 278 (2000), 2039-2049. doi: 10.1152/ajpheart.2000.278.6.H2039
[6]	D. Karabulut, F. Ortes, Y. Z. Arslan, M. A. Adli, Comparative evaluation of EMG signal features for myoelectric controlled human arm prosthetics, Biocybern. Biomed. Eng., 37 (2017), 326-335. doi: 10.1016/j.bbe.2017.03.001
[7]	A. Phinyomark, F. Quaine, S. Charbonnier, C. Serviere, F. Tarpin-Bernard, Y. Laurillau, EMG feature evaluation for improving myoelectric pattern recognition robustness, Expert Syst. Appl., 40 (2013), 4832-4840.
[8]	J. J. Wang, F. R. Gao, Y. Sun, Z. Z. Luo, Non-uniform characteristics and its recognition effects for walking gait based on sEMG, Chinese J. Sensor. Actuat., 29 (2016), 384-389.
[9]	P. Sbriccoli, I. Bazzucchi, A. Rosponi, M. Bernardi, G. D. Vito, F. Felici, Amplitude and spectral characteristics of biceps brachii sEMG depend upon speed of isometric force generation, J. Electromyogr. Kinesiol., 13 (2003), 139-147. doi: 10.1016/S1050-6411(02)00098-6
[10]	S. Pancholi, A. M. Joshi, Portable EMG data acquisition module for upper limb prosthesis application, IEEE Sens. J., 18 (2018), 3436-3443. doi: 10.1109/JSEN.2018.2809458
[11]	L. Zhang, G. Liu, B. Han, Z. Wang, T. Zhang, sEMG based human motion intention recognition, J. Robot., 2019 (2019), 3679174.
[12]	C. Cortes, V. N. Vapnik, Support-vector networks, Mach. Learn., 20 (1995), 273-297.
[13]	G. B. Huang, Q. Y. Zhu, C. K. Siew, Extreme learning machine: a new learning scheme of feedforward neural networks, IEEE IJCNN, (2004), 985-990.
[14]	I. S. Dhindsa, R. Agarwal, H. S. Ryait, Performance evaluation of various classifiers for predicting knee angle from electromyography signals, Expert Syst., 36 (2019), 1-14.
[15]	Y. S. Wu, S. Liang, L. Zhang, Z. Q. Chai, C. C. Cao, S. W. Wang, Gesture recognition method based on a single-channel sEMG envelope signal, EURASIP J. Wirel. Commun. Netw., 35 (2018), 1-8.
[16]	C. Castellini, P. Smagt, Surface EMG in advanced hand prosthetics, Biol. Cybern., 100 (2009), 35-47. doi: 10.1007/s00422-008-0278-1
[17]	Y. X. Kuang, Q. Wu, J. K. Shao, J. F. Wu, X. H. Wu, Extreme learning machine classification method for lower limb movement recognition, Cluster Comput., 20 (2017), 3051-3059. doi: 10.1007/s10586-017-0985-2
[18]	X. G. Xi, M. Y. Tang, S. M. Miran, Z. Z. Luo, Evaluation of feature extraction and recognition for activity monitoring and fall detection based on wearable sEMG sensors, Sensors, 17 (2017), 1229.
[19]	Y. X. Deng, F. R. Gao, H. H. Chen, Angle estimation for knee joint movement based on PCA-RELM algorithm, Symmetry, 12 (2020), 130.
[20]	F. R. Gao, T. X. Tian, T. Yao, Q. Z. Zhang, Human gait recognition based on multiple feature combination and parameter optimization algorithms, Comput. Intell. Neurosci., 2021 (2021), 6693206.
[21]	H. T. Vu, D. B. Dong, H. L. Cao, T. Verstraten, D. Lefeber, B. Vanderborght, et al., A review of gait phase detection algorithms for lower limb prostheses, Sensors, 2019 (2020), 3972.
[22]	Y. L. Cun, Y. Bengio, G. Hinton, Deep learning, Nature, 521 (2015), 436-444.
[23]	C. Morbidoni, A. Cucchiarelli, S. Fioretti, F. D. Nardo, A deep learning approach to EMG-based classification of gait phases during level ground walking, Electronics, 8 (2019), 894.
[24]	J. C. Chen, X. D. Zhang, Y. Cheng, N. Xi, Surface EMG based continuous estimation of human lower limb joint angles by using deep belief networks, Biomed. Signal Process. Control, 40 (2018), 335-342. doi: 10.1016/j.bspc.2017.10.002
[25]	M. Simao, N. Mendes, O. Gibaru, P. Neto, A review on electromyography decoding and pattern recognition for human-machine interaction, IEEE Access, 7 (2019), 39564-39582. doi: 10.1109/ACCESS.2019.2906584
[26]	A. K. Mukhopadhyay, S. Samui, An experimental study on upper limb position invariant EMG signal classification based on deep neural network, Biomed. Signal Process. Control, 55 (2020), 101669.
[27]	J. H. Kim, G. S. Hong, B. G. Kim, D. P. Dogra, Deepgesture: deep learning-based gesture recognition scheme using motion sensors, Displays, 55 (2018), 38-45. doi: 10.1016/j.displa.2018.08.001
[28]	D. Jeong, B. G. Kim, S. Y. Dong, Deep joint spatiotemporal network (DJSTN) for efficient facial expression recognition, Sensors, 20 (2020), 1963.
[29]	J. H. Kim, B. G. Kima, P. P. Roy, D. M. Jeong, Efficient facial expression recognition algorithm based on hierarchical deep neural network structure, IEEE Access, 7 (2019), 41273-41285. doi: 10.1109/ACCESS.2019.2907327
[30]	R. N. Khushaba, A. A. Ani, A. A. Timemy, A. A. Jumaily, A fusion of time-domain descriptors for improved myoelectric hand control, IEEE SSCI, (2016), 1-6.
[31]	A. Gharehbaghi, M. Linden, A deep machine learning method for classifying cyclic time series of biological signals using time-growing neural network, IEEE Trans. Neur. Netw. Learn. Syst., 29 (2018), 4102-4115. doi: 10.1109/TNNLS.2017.2754294
[32]	G. D. Guo, N. Zhang, A survey on deep learning based face recognition, Comput. Vis. Image Underst., 189 (2019), 1-37.
[33]	C. J. Li, J. Ren, H. Q. Huang, B. Wang, Y. F. Zhu, H. S. Hu, PCA and deep learning based myoelectric grasping control of a prosthetic hand, Biomed. Eng. Online, 17 (2018), 107.
[34]	U. C. Allard, C. L. Fall, A. Drouin, A. C. Lecours, C. Gosselin, K. Glette, et al., Deep learning for electromyographic hand gesture signal classification using transfer learning, IEEE Trans. Neur. Syst. Rehabil. Eng., 27(2019), 760-771.
[35]	A. Phinyomark, P. Phukpattaranont, C. Limsakul, Feature reduction and selection for EMG signal classification, Expert Syst. Appl., 39 (2012), 7420-7431. doi: 10.1016/j.eswa.2012.01.102
[36]	M. Zardoshti-Kermani, B. C. Wheeler, K. Badie, R. M. Hashemi, EMG feature evaluation for movement control of upper extremity prostheses, IEEE Trans. Rehabil. Eng., 3 (1995), 324-333. doi: 10.1109/86.481972
[37]	R. Boostani, M. H. Moradi, Evaluation of the forearm EMG signal features for the control of a prosthetic hand, Physiol. Meas., 24 (2003), 309-319. doi: 10.1088/0967-3334/24/2/307
[38]	Y. Y. Cao, F. R. Gao, L. Yu, Q. S. She, Gait recognition based on EMG information with multiple features, IFIP Adv. Inf. Commun. Technol., (2018), 402-411.
[39]	J. H. Li, G. F. Li, Y. Sun, G. Z. Jiang, B. Tao, S. Xu, Hand motions recognition based on sEMG nonlinear feature and time domain feature fusion, Int. J. Innovat. Comput. Appl., 10 (2019), 43-50. doi: 10.1504/IJICA.2019.100510
[40]	Y. Narayan, Hb vsEMG signal classification with time domain and Frequency domain features using LDA and ANN classifier, Mater. Today, 37 (2021), 3226-3230.
[41]	G. E. Hinton, R. R. Salakhutdinov, Reducing the dimensionality of data with neural networks, Science, 313 (2006), 504-507. doi: 10.1126/science.1127647
[42]	R. L. Ortolan, R. N. Mori, R. R. Pereira, C. M. N. Cabral, J. C. Pereira, A. Cliquet, Evaluation of adaptive/nonadaptive filtering and wavelet transform techniques for noise reduction in EMG mobile acquisition equipment, IEEE Trans. Neur. Syst. Rehabil. Eng., 11 (2003), 60-69. doi: 10.1109/TNSRE.2003.810432
[43]	S. G. Chang, B. Yu, M. Vetterli, Adaptive wavelet thresholding for image denoising and compression, IEEE Trans. Image Process., 9 (2000), 1532-1546. doi: 10.1109/83.862633
[44]	A. Fathi, A. R. Naghsh-Nilchi, Efficient image denoising method based on a new adaptive wavelet packet thresholding function, IEEE Trans. Image Process., 21 (2012), 3981-3990. doi: 10.1109/TIP.2012.2200491
[45]	G. E. Hinton, S. Osindero, Y. W. Teh, A fast learning algorithm for deep belief nets, Neur. Comput., 18 (2006), 1527-1554. doi: 10.1162/neco.2006.18.7.1527
[46]	C. Szegedy, W. Liu, Y. Q. Jia, P. Sermanet, S. Reed, D. Anguelov, et al., Going deeper with convolutions, IEEE CVPR, Boston, MA, USA, (2015), 1-9.
[47]	R. Kohavi, A study of cross-validation and bootstrap for accuracy estimation and model selection, in Proceedings of the 14th International Joint Conference on Artificial intelligence, 2 (1995), 1137-1143.

Reader Comments

Your name:*

Email:*
© 2021 the Author(s), licensee AIMS Press. This is an open access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0)