ANALYSIS AND DATA PROCESSING SYSTEMS
Print ISSN: 2782-2001
Online ISSN: 2782-215X
The problem of preprocessing initial data in order to isolate informative features of the EMG signal in the time domain is solved in order to classify mimic movements. The extracted features are processed by the artificial neural network (ANN) classifier on the basis of radial-basis functions (RBS). To increase the effectiveness of ANN training, it was suggested to use the method of biofeedback (BF), which allows improving the accuracy of the classifier due to lesser variability of an input signal for various gestures. The results of the experiment on the study of the effectiveness of the mimic gesture classifier operating in real time are presented.
A group of ten volunteers was involved in the experiment to obtain a sample for training the classifier. They experimentally estimated the efficiency of using the characteristics of the classifier of six types of EMG features computed in the time domain as an input vector. As a result of the comparison, a high informativity of such an EMG attribute as a signal envelope calculated by means of the Hilbert transform with subsequent averaging over peak values and root-mean-square deviation is proved. As a tool for pre-processing initial data for feature extraction, we can recommend the construction of an envelope with averaging over peak values for 10 signal readings (at a sampling frequency of 1 kHz) as an input feature vector. The error of recognizing gestures with the use of the proposed classifier in real time was no more than 4.8%, which is an acceptable level for using the classifier as part of control systems for household devices.
Chernov Nikolay N.
1. Uroven' invalidizatsii v Rossiiskoi Federatsii [The level of disability in the Russian Federation]. Federal'naya sluzhba gosudarstvennoi statistiki [Federal State Statistics Service]: website. Available at: http://www.gks.ru/wps/wcm/connect/rosstat_main/rosstat/ru/statistics/population/disabilities/# (accessed 22.05.2018).
2. Ushiba J., Soekadar S. Brain-machine interfaces for rehabilitation of poststroke hemiplegia. Progress in Brain Research, 2016, vol. 228, pp. 163–183.
3. Georgi M., Amma C., Schultz T. Recognizing hand and finger gestures with IMU based motion and EMG based muscle activity sensing. Proceedings of the International Conference on Bio-inspired Systems and Signal Processing (BIOSIGNALS-2015), Lisbon, Portugal, 2015, vol. 4, pp. 99–108.
4. Prajwal P., Ayan B., Sandeep K. SCEPTRE: a pervasive, non-invasive, and programmable gesture recognition technology. Proceedings of the 21st International Conference on Intelligent User InterfacesSonoma, CA, USA, 2016, pp. 282–293.
5. Pradeep Reddy C.H. An overview on human gesture recognition. International Journal Of Pharmacy & Technology 2016, vol. 8, no. 2, pp. 12037–12045.
6. Naseer N., Determining optimal feature-combination for LDA classification of functional near-infrared spectroscopy signals in brain-computer interface application. Frontiers in Human Neuroscience 2016, vol. 10, art. 237. Available at: https://doi.org/10.3389/fnhum.2016.00237 (accessed 22.05.2018).
7. Redlarski G., Gradolewski D., Palkowski A. A system for heart sounds classification. PLOS ONE, 2014, vol. 9 (11). Available at: https://doi.org/10.1371/journal.pone.0112673. (accessed 22.05.2018).
8. Gesture recognition using real time EMG. 2015 International Conference on Innovations in Information, Embedded and Communication Systems (ICIIECS), Coimbatore, India, 2015. Available at: https://doi.org/10.1109/ICIIECS.2015.7193196 (accessed 22.05.2018).
9. Nikolaev S.G. Praktikum po klinicheskoi elektromiografii [Practical work on clinical electromyography]. Ivanovo, IGMA Publ., 2003. 264 p.
10. Shevtsov V.I., Skripnikov A.A., Shein A.P. Primenenie biologicheskoi obratnoi svyazi po elektromiogramme v kompleksnom lechenii bol'nykh s tsentral'nymi gemiparezami (obzor literatury) [The use of biological feedback by electromyogram in complex treatment of patients with central hemiparesis (review of literature)]. Genii ortopedii – The genius of orthopedics 2007, iss. 1, pp. 142–147.
11. Komantsev V.N. Metodicheskie osnovy klinicheskoi elektroneiromiografii: rukovodstvo dlya vrachei [Methodical bases of clinical electroneuromyography. Guidelines for doctors]. St. Petersburg, 2006. 362 p.
12. Budko R.Yu., Starchenko I.B. Sozdanie klassifikatora mimicheskikh dvizhenii na osnove analiza elektromiogrammy [Creation of the facial gestures сlassifier based on the electromyogram analysis]. Trudy SPIIRAN – SPIIRAS Proceedings, 2016, iss. 46, pp. 76–89.
13. Budko R., Starchenko I., Budko A. Preprocessing data for facial gestures classifier on the basis of the neural network analysis of biopotentials muscle signals. ICR 2016: Interactive Collaborative Robotics, Proceedings Cham, Springer International Publishing, 2016, pp. 163–171.
14. Anetha K, Rejina J. Hand talk a sign language recognition based on accelerometer and SEMG data. International Journal of Innovative Research in Computer and Communication Engineering 2014, vol. 2, Special issue 3, pp. 206–215.
15. Li H., Zhao G., Zhou Y., Chen X., Ji Z., Wang L. Relationship of EMG/SMG features and muscle strength level: an exploratory study on tibialis anterior muscles during plantar-flexion among hemiplegia patients. BioMedical Engineering OnLine 2014, vol. 13. Available at: https://biomedical-engineering-online.biomedcentral.com/articles/10.1186/1475-925X-13-5 (accessed 22.05.2018).
16. Sushkova O.S., Korolev M.S., Gabova A.V., Illarioshkin S.N., Karabanov A.V., Obukhov K.Yu., Obukhov Yu.V. Sovmestnyi analiz signalov elektroentsefalogramm, elektromiogramm i mekhanicheskogo tremora pri bolezni Parkinsona v rannei stadii [Joint analysis of signals of electroencephalograms, electromyograms and mechanical tremor in Parkinson's disease at an early stage]. Zhurnal radioelektroniki – Journal of Radioelectronics 2014, vol. 5. Available at: http://jre.cplire.ru/iso/may14/12/text.pdf (accessed 17.02.17).
17. Englehart K., Hudgins B. A robust, real-time control scheme for multifunction myoelectric control. IEEE Translocation Biomedical Engineering 2003, vol. 50, pp. 848–854.
18. Pattern recognition and machine learning. New York, Springer, 2006. 738 p.
This work was supported by “UMNIK” awards №11689ГУ/2017 to R.Yu. Budko.
Budko R.Yu., Chernov N.N, Budko A.Yu. Raspoznavanie myshechnykh usilii po signalu litsevoi elektromiogrammy v rezhime real'nogo vremeni [Recognition of facial movements by facial electromyogram signals in real time]. Nauchnyi vestnik Novosibirskogo gosudarstvennogo tekhnicheskogo universiteta – Science bulletin of the Novosibirsk state technical university, 2018, no. 2 (71), pp. 59–74. doi: 10.17212/1814-1196-2018-2-59-74.
