Journal of Internet of Things and Convergence (사물인터넷융복합논문지)

The Korea Internet of Things Society (한국사물인터넷학회)
권호 표지
DOI QR코드

DOI QR Code

Recognition of hand gestures with different prior postures using EMG signals

사전 자세에 따른 근전도 기반 손 제스처 인식

  • Hyun-Tae Choi (Artificial Intelligence and Convergence Department, Pukyong National University) ;
  • Deok-Hwa Kim (Department of Computer Engineering, Pukyong National University) ;
  • Won-Du Chang (Division of Computer Engineering and Artificial Intelligence, Pukyong National University)
  • 최현태 (부경대학교 인공지능융합학과) ;
  • 김덕화 (부경대학교 컴퓨터공학과) ;
  • 장원두 (부경대학교 컴퓨터.인공지능공학부)
  • Received : 2023.09.20
  • Accepted : 2023.11.20
  • Published : 2023.12.31
Download PDF
⟨ Previous Next ⟩

Abstract

Hand gesture recognition is an essential technology for the people who have difficulties using spoken language to communicate. Electromyogram (EMG), which is often utilized for hand gesture recognition, is expected to have difficulties in hand gesture recognition because its people's movements varies depending on prior postures, but the study on this subject is rare. In this study, we conducted tests to confirm if the prior postures affect on the accuracy of gesture recognition. Data were recorded from 20 subjects with different prior postures. We achieved average accuracies of 89.6% and 52.65% when the prior states between the training and test data were unique and different, respectively. The accuracy was increased when both prior states were considered, which confirmed the need to consider a variety of prior states in hand gesture recognition with EMG.

손 제스처의 인식은 구어 사용이 어려운 사람들의 의사소통을 위한 중요한 기술이다. 제스처 인식에 널리 사용되는 근전도 신호는 사전 자세에 따라 동작이 달라지기 때문에 제스처 인식의 어려움이 있을 것으로 예상되지만, 이에 관한 연구는 찾기 어렵다. 본 연구에서는 사전 자세에 따른 제스처 인식 성능의 변화를 분석하였다. 이를 위해 총 20명의 피험자에게서 사전 자세를 가지는 동작에 대한 근전도 신호를 측정하고, 제스처 인식을 실험하였다. 그 결과, 학습 및 테스트 데이터 간 사전 상태가 단일한 경우에는 평균 89.6%의 정확도를, 상이한 경우에는 평균 52.65%의 정확도를 보였다. 반면, 사전 자세를 모두 고려한 경우에는 정확도가 다시 회복됨을 발견하였다. 이를 통해 본 연구에서는 근전도를 활용하는 손 제스처 인식시에 사전 자세가 다양하게 고려하여야 함을 실험적으로 확인하였다.

Keywords

Acknowledgement

본 논문은 부경대학교 자율창의학술연구비(2022년)에 의하여 연구되었음.

References

  1. Korean Statistical Information Service[Internet], https://kosis.kr/statHtml/statHtml.do?orgId=117&tblId=DT_11761_N003
  2. R.L.Weerasinghe and G.U.Ganegoda, "A Comprehensive Review on Vision-based Sign Language Detection and Recognition," 2022 International Research Conference on Smart Computing and Systems Engineering, Vol.5, pp.88-95, 2022.
  3. S.A.M.A.S.Senanayaka, R.A.D.B.S.Perera, W.Rankothge, S.S.Usgalhewa, H.D.Hettihewa and P.K.W.Abeygunawardhana, "Continuous American Sign Language Recognition Using Computer Vision and Deep Learning Technologies," 2022 IEEE Region 10 Symposium, pp.1-6, 2022.
  4. I.H.Kim and I.H.Jung, "A Study on Korea Sign Language Motion Recognition Using OpenPose Based on Deep Learning," Journal of Digital Contents Society, Vol.22, No.4, pp.681-687, 2021. https://doi.org/10.9728/dcs.2021.22.4.681
  5. S.Shin, Y.Baek, J.Lee, Y.Eun and S.H.Son, "Korean Sign Language Recognition Using EMG and IMU Sensors Based on Group-Dependent NN Models," 2017 IEEE Symposium Series on Computational Intelligence, pp.1-7, 2017.
  6. D.G.Yuk and J.W.Sohn, "Hand Gesture Recognition Using Surface Electromyogram," Korean Soc. Noise Vib. Eng., Vol.28, No.6, pp.670-676, 2018. https://doi.org/10.5050/KSNVE.2018.28.6.670
  7. C.Savur and F.Sahin, "Real-Time American Sign Language Recognition System Using Surface EMG Signal," 2015 IEEE 14th International Conference on Machine Learning and Applications, pp.497-502, 2015.
  8. A.B.H.Amor, O.El Ghoul and M.Jemni, "A deep learning based approach for Arabic Sign Language Alphabet Recognition Using Electromyographic Signals," 2021 8th International Conference on ICT & Accessibility, pp.1-4, 2021.
  9. S.Yuan, Y.Wang, X.Wang, H.Deng, S.Sun, H.Wang, P.Huang and G.Li, "Chinese Sign Language Alphabet Recognition Based on Random Forest Algorithm," 2020 IEEE International Workshop on Metrology for Industry 4.0 & IoT, pp.340-344, 2022.
  10. J.Qi, G.Jiang, G.Li, Y.Sun, and B.Tao, "Surface EMG Hand Gesture Recognition System Based on PCA and GRNN," Neural Computing and Applications, Vol.32, pp.6343-6351, 2020. https://doi.org/10.1007/s00521-019-04142-8
  11. M.A.Ozdemir, D.H.Kisa, O.Guren, A.Onan, and A.Akan, "EMG Based Hand Gesture Recognition Using Deep Learning," In 2020 Medical Technologies Congress, pp.1-4, 2020.
  12. X.Chen, Y.Li, R.Hu, X.Zhang, and X.Chen, "Hand Gesture Recognition Based on Surface Electromyography Using Convolutional Neural Network with Transfer Learning Method," IEEE Journal of Biomedical and Health Informatics, Vol.25, No.4, pp.1292-1304, 2020.
  13. J.J.Park and C.K.Kwon, "Study on Forearm Muscles and Electrode Placements for CNN Based Korean Finger Number Gesture Recognition using sEMG Signals," Journal of the Korea Academia-Industrial Cooperation Society Vol.19, No.8, pp.260-267, 2018. https://doi.org/10.5762/KAIS.2018.19.8.260
  14. Myoware muscle sensor[Internet], https://www.digikey.com/en/maker/projects/myoware-muscle-sensor-kit/39ecba5502ad4c59ad0f7eca25b6e338
  15. M.Jordanic, M.Rojas-Martinez, M.Mananas, J.F.Alonso and H.R.Marateb, "A Novel Spatial Feature for The Identification of Motor Tasks Using High-Density Electromyography," Sensors, Vol.17, No.7, pp.1597, 2017.
  16. S.-H.Kim, H.-R.Shin, Y.Han, W.-D.Chang, Y.-J.Min, "Gesture Recognition of Writing Numbers Using Pressure Sensor Based on CNN-LSTM Combination Model, Journal of The Institute of Electronics and Information Engineers," 2023. (In press)
  17. T.T.Um, F.M.Pfister, D.Pichler, S.Endo, M.Lang, S.Hirche, U.Fietzek and D.Kulic, "Data Augmentation of Wearable Sensor Data for Parkinson's Disease Monitoring Using Convolutional Neural Networks," Proceedings of the 19th ACM International Conference on Multimodal Interaction. pp.216-220, 2017.
  18. A.L.Guennec, S.Malinowski, and R.Tavenard. "Data Augmentation for Time Series Classification Using Convolutional Neural Networks," ECML/PKDD Workshop on Advanced Analytics and Learning on Temporal Data. 2016.