Human posture recognition using android smartphone and artificial neural network
Current smartphones are equipped with various sensors, which can be used for research and data collection purposes. This papers presents an approach to use the gyroscope sensor present in many smartphones to perform gesture recognition. Two phones were strapped onto the subject body. Gyroscope readi...
| Published in: | Proceedings - 2015 6th IEEE Control and System Graduate Research Colloquium, ICSGRC 2015 |
|---|---|
| Main Author: | |
| Format: | Conference paper |
| Language: | English |
| Published: |
Institute of Electrical and Electronics Engineers Inc.
2016
|
| Online Access: | https://www.scopus.com/inward/record.uri?eid=2-s2.0-84964498740&doi=10.1109%2fICSGRC.2015.7412477&partnerID=40&md5=b9fe8041a4fb4339bb8ee1f498aa57d6 |
| Summary: | Current smartphones are equipped with various sensors, which can be used for research and data collection purposes. This papers presents an approach to use the gyroscope sensor present in many smartphones to perform gesture recognition. Two phones were strapped onto the subject body. Gyroscope readings were obtained during several gestures. The gyroscope readings were sent to MATLAB using the SensorUDP application installed on the phone. A total of 125 readings of 4 gestures were collected from 4 subjects and fed to a Multi-Layer Perceptron (MLP) classifier. Tests were performed to determine the optimal threshold and number of hidden units, respectively. The best classifier produced 99.69% accuracy. © 2015 IEEE. |
|---|---|
| ISSN: | |
| DOI: | 10.1109/ICSGRC.2015.7412477 |