근전도 신호 기반 Multi-Stream CNN 및 경험적 모드 분해를 이용한 사용자 인식

다국어 초록 (Multilingual Abstract)

The Electromyogram(EMG) signal is a biosignal generated during contraction of the skeletal muscle, and different signal waveforms are generated according to the performed motion. Accordingly, it is possible to solve the problem that the registration information of the existing personal recognition method cannot be changed. In this paper, we propose a personal recognition method using EMG signal-based Empirical Mode Decomposition(EMD) and multi-stream CNN. The proposed method decomposes the EMG signal into an Intrinsic Mode Function(IMF) using EMD for increase feature data. The decomposed IMF 1-4 is used as input data to the designed multi-stream CNN to perform personal recognition. As a result of the experiment using Ninapro DB2, which is benchmarking EMG data, the method using the multi-stream CNN and the IMF showed an accuracy of 98.3%, which was 1% higher than the existing study using CNN.

참고문헌 (Reference)

1 김경래 ; 최홍락 ; 김경석, "측면형 지정맥 인식기 설계 및 구현" 한국인터넷방송통신학회 21 (21): 159-168, 2021

2 이진욱 ; 이현욱 ; 유철상, "이변량 웨이블릿 분석을 위한 모 웨이블릿 선정" 한국수자원학회 52 (52): 905-916, 2019

3 김진수 ; 김성혁 ; 반성범, "사전 학습된 네트워크 모델을 이용한 심전도 신호 기반 개인 식별 성능 분석" 한국정보기술학회 18 (18): 107-114, 2020

4 김경래 ; 최홍락 ; 김경석, "고성능 CNN 기반 지정맥 인증 시스템 구현" 한국인터넷방송통신학회 21 (21): 197-202, 2021

5 S. Morikawa, "Personal authentication by lips EMG using dry electrode and CNN" 180-183, 2018

6 X. Song, "Motion recognition of the bilateral upper-limb rehabilitation using sEMG based on ensemble EMD" 1637-1642, 2014

7 C. Sapsanis, "Improving EMG based classification of basic hand movements using EMD" 5754-5757, 2013

8 M. Atzori, "Electromyography data for non-invasive naturally-controlled robotic hand prostheses" 1 : 1-13, 2014

9 S. Aziz, "Electricity theft detection using empirical mode decomposition and k-nearest neighbors" 1-5, 2020

10 김상호 ; 유재환 ; 이병현 ; 김덕환, "EMG 신호 기반 Artificial Neural Network을이용한 사용자 인식" 대한전자공학회 53 (53): 142-148, 2016

11 M. Karuna, "EMG signal analysis using intrinsic mode functions to discriminate upper limb movement" 1-3, 2020

12 D. H. Kisa, "EMG based hand gesture classification using empirical mode decomposition time-series and deep learning" 1-4, 2020

13 김재명 ; 최규호 ; 김진수 ; 반성범, "CNN 기반 근전도의 2D 스펙트로그램 이미지를 이용한 사용자 인식" 한국정보기술학회 19 (19): 107-117, 2021

14 M. R. Lee, "Biometric personal identification based on gait analysis using surface EMG signals" 318-321, 2017

15 M. U. Khan, "Biometric authentication based on EMG signals of speech" 1-5, 2020

16 S. H. Shin, "A study on EMG signal acquisition modules and artificial neural networks for personal authentication" 1458-1459, 2018

17 L. Lu, "A study of personal recognition method based on EMG signal" 14 (14): 681-691, 2020

18 M. G. Kim, ""Deep learning based on 1-D ensemble networks using ECG for real-time user recognition" 15 (15): 5656-5663, 2019