Real-time recognition of arm motion using artificial neural network multi-perceptron with arduino one microcontroller and EKG/EMG shield sensor

Luis A. Caro, Camilo Silva, Billy Peralta, Oriel A. Herrera, Sergio Barrientos

Resultado de la investigación: Conference contribution

Resumen

Currently, human-computer interfaces have a number of useful applications for people. The use of electromyographic signals (EMG) has shown to be effective for human-computer interfaces. The classification of patterns based on EMG signals has been successfully applied in various tasks such as motion detection to control of video games. An alternative to increasing access to these applications is the use of low-cost hardware to sample the EMG signals considering a real-time response. This paper presents a methodology for recognizing patterns of EMG signals given by arm movements in real time. Our proposal is based on an artificial Neural Network, Multilayer Perceptron, where the EMG signals are processed by a set of signal processing techniques. The hardware used for obtaining the signal is based on Ag/AgCl connected to the EKG/EMG-Shield plate mounted on a Arduino One R3 card which is used to control a video game. The implemented application achieves an accuracy above 90% using less than 0.2 s for recognition of actions in time of testing. Our methodology is shown to predict different movements of the human arm reliably, at a low cost and in real time.

Idioma originalEnglish
Título de la publicación alojadaAmbient Intelligence for Health - 1st International Conference, AmIHEALTH 2015, Proceedings
EditoresVladimir Villarreal, José Bravo, Ramón Hervás
EditorialSpringer Verlag
Páginas3-14
Número de páginas12
ISBN (versión impresa)9783319265070
DOI
EstadoPublished - 1 ene 2015
Evento1st International Conference on Ambient Intelligence for Health, AmIHEALTH 2015 - Puerto Varas, Chile
Duración: 1 dic 20154 dic 2015

Serie de la publicación

NombreLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volumen9456
ISSN (versión impresa)0302-9743
ISSN (versión digital)1611-3349

Conference

Conference1st International Conference on Ambient Intelligence for Health, AmIHEALTH 2015
PaísChile
CiudadPuerto Varas
Período1/12/154/12/15

Huella dactilar

Microcontroller
Microcontrollers
Perceptron
Electrocardiography
Artificial Neural Network
Neural networks
Real-time
Sensor
Interfaces (computer)
Motion
Sensors
Hardware
Multilayer neural networks
Costs
Signal processing
Human-computer Interface
Video Games
Testing
Motion Detection
Methodology

ASJC Scopus subject areas

  • Theoretical Computer Science
  • Computer Science(all)

Citar esto

Caro, L. A., Silva, C., Peralta, B., Herrera, O. A., & Barrientos, S. (2015). Real-time recognition of arm motion using artificial neural network multi-perceptron with arduino one microcontroller and EKG/EMG shield sensor. En V. Villarreal, J. Bravo, & R. Hervás (Eds.), Ambient Intelligence for Health - 1st International Conference, AmIHEALTH 2015, Proceedings (pp. 3-14). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9456). Springer Verlag. https://doi.org/10.1007/978-3-319-26508-7_1
Caro, Luis A. ; Silva, Camilo ; Peralta, Billy ; Herrera, Oriel A. ; Barrientos, Sergio. / Real-time recognition of arm motion using artificial neural network multi-perceptron with arduino one microcontroller and EKG/EMG shield sensor. Ambient Intelligence for Health - 1st International Conference, AmIHEALTH 2015, Proceedings. editor / Vladimir Villarreal ; José Bravo ; Ramón Hervás. Springer Verlag, 2015. pp. 3-14 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{d4b52b4ab66b44c0a14bb729da4fb3a3,
title = "Real-time recognition of arm motion using artificial neural network multi-perceptron with arduino one microcontroller and EKG/EMG shield sensor",
abstract = "Currently, human-computer interfaces have a number of useful applications for people. The use of electromyographic signals (EMG) has shown to be effective for human-computer interfaces. The classification of patterns based on EMG signals has been successfully applied in various tasks such as motion detection to control of video games. An alternative to increasing access to these applications is the use of low-cost hardware to sample the EMG signals considering a real-time response. This paper presents a methodology for recognizing patterns of EMG signals given by arm movements in real time. Our proposal is based on an artificial Neural Network, Multilayer Perceptron, where the EMG signals are processed by a set of signal processing techniques. The hardware used for obtaining the signal is based on Ag/AgCl connected to the EKG/EMG-Shield plate mounted on a Arduino One R3 card which is used to control a video game. The implemented application achieves an accuracy above 90{\%} using less than 0.2 s for recognition of actions in time of testing. Our methodology is shown to predict different movements of the human arm reliably, at a low cost and in real time.",
keywords = "Action recognition, Arduino, Microcontrollers, Neural networks",
author = "Caro, {Luis A.} and Camilo Silva and Billy Peralta and Herrera, {Oriel A.} and Sergio Barrientos",
year = "2015",
month = "1",
day = "1",
doi = "10.1007/978-3-319-26508-7_1",
language = "English",
isbn = "9783319265070",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "3--14",
editor = "Vladimir Villarreal and Jos{\'e} Bravo and Ram{\'o}n Herv{\'a}s",
booktitle = "Ambient Intelligence for Health - 1st International Conference, AmIHEALTH 2015, Proceedings",
address = "Germany",

}

Caro, LA, Silva, C, Peralta, B, Herrera, OA & Barrientos, S 2015, Real-time recognition of arm motion using artificial neural network multi-perceptron with arduino one microcontroller and EKG/EMG shield sensor. En V Villarreal, J Bravo & R Hervás (eds.), Ambient Intelligence for Health - 1st International Conference, AmIHEALTH 2015, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 9456, Springer Verlag, pp. 3-14, 1st International Conference on Ambient Intelligence for Health, AmIHEALTH 2015, Puerto Varas, Chile, 1/12/15. https://doi.org/10.1007/978-3-319-26508-7_1

Real-time recognition of arm motion using artificial neural network multi-perceptron with arduino one microcontroller and EKG/EMG shield sensor. / Caro, Luis A.; Silva, Camilo; Peralta, Billy; Herrera, Oriel A.; Barrientos, Sergio.

Ambient Intelligence for Health - 1st International Conference, AmIHEALTH 2015, Proceedings. ed. / Vladimir Villarreal; José Bravo; Ramón Hervás. Springer Verlag, 2015. p. 3-14 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 9456).

Resultado de la investigación: Conference contribution

TY - GEN

T1 - Real-time recognition of arm motion using artificial neural network multi-perceptron with arduino one microcontroller and EKG/EMG shield sensor

AU - Caro, Luis A.

AU - Silva, Camilo

AU - Peralta, Billy

AU - Herrera, Oriel A.

AU - Barrientos, Sergio

PY - 2015/1/1

Y1 - 2015/1/1

N2 - Currently, human-computer interfaces have a number of useful applications for people. The use of electromyographic signals (EMG) has shown to be effective for human-computer interfaces. The classification of patterns based on EMG signals has been successfully applied in various tasks such as motion detection to control of video games. An alternative to increasing access to these applications is the use of low-cost hardware to sample the EMG signals considering a real-time response. This paper presents a methodology for recognizing patterns of EMG signals given by arm movements in real time. Our proposal is based on an artificial Neural Network, Multilayer Perceptron, where the EMG signals are processed by a set of signal processing techniques. The hardware used for obtaining the signal is based on Ag/AgCl connected to the EKG/EMG-Shield plate mounted on a Arduino One R3 card which is used to control a video game. The implemented application achieves an accuracy above 90% using less than 0.2 s for recognition of actions in time of testing. Our methodology is shown to predict different movements of the human arm reliably, at a low cost and in real time.

AB - Currently, human-computer interfaces have a number of useful applications for people. The use of electromyographic signals (EMG) has shown to be effective for human-computer interfaces. The classification of patterns based on EMG signals has been successfully applied in various tasks such as motion detection to control of video games. An alternative to increasing access to these applications is the use of low-cost hardware to sample the EMG signals considering a real-time response. This paper presents a methodology for recognizing patterns of EMG signals given by arm movements in real time. Our proposal is based on an artificial Neural Network, Multilayer Perceptron, where the EMG signals are processed by a set of signal processing techniques. The hardware used for obtaining the signal is based on Ag/AgCl connected to the EKG/EMG-Shield plate mounted on a Arduino One R3 card which is used to control a video game. The implemented application achieves an accuracy above 90% using less than 0.2 s for recognition of actions in time of testing. Our methodology is shown to predict different movements of the human arm reliably, at a low cost and in real time.

KW - Action recognition

KW - Arduino

KW - Microcontrollers

KW - Neural networks

UR - http://www.scopus.com/inward/record.url?scp=84954158366&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-26508-7_1

DO - 10.1007/978-3-319-26508-7_1

M3 - Conference contribution

SN - 9783319265070

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 3

EP - 14

BT - Ambient Intelligence for Health - 1st International Conference, AmIHEALTH 2015, Proceedings

A2 - Villarreal, Vladimir

A2 - Bravo, José

A2 - Hervás, Ramón

PB - Springer Verlag

ER -

Caro LA, Silva C, Peralta B, Herrera OA, Barrientos S. Real-time recognition of arm motion using artificial neural network multi-perceptron with arduino one microcontroller and EKG/EMG shield sensor. En Villarreal V, Bravo J, Hervás R, editores, Ambient Intelligence for Health - 1st International Conference, AmIHEALTH 2015, Proceedings. Springer Verlag. 2015. p. 3-14. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-319-26508-7_1