Optical myography sensor for gesture recognition
- Resource Type
- Conference
- Authors
- Wu, Yu Tzu; Fujiwara, Eric; Suzuki, Carlos Kenichi
- Source
- 2018 IEEE 15th International Workshop on Advanced Motion Control (AMC) Advanced Motion Control (AMC), 2018 IEEE 15th International Workshop on. :348-352 Mar, 2018
- Subject
- Power, Energy and Industry Applications
Robotics and Control Systems
Signal Processing and Analysis
Muscles
Thumb
Monitoring
Skin
Computer vision
Image color analysis
Cameras
Optical myography
computer vision sensor
gesture recognition
human-computer interaction
- Language
- ISSN
- 1943-6580
In this work, an optical myography system is demonstrated as an innovative and promising alternative to recover gesture information. During the experiments, a total of eight different postures were defined to evaluate the sensor performance. A classifier based on artificial neural network was trained and validated, showing an average precision of ~ 92% and accuracy of ~ 98%.