Research Article
Motion Recognition Uses Accelerometer and Gyroscope Sensors Using Learning Vector Quantization Method
@INPROCEEDINGS{10.4108/eai.7-11-2023.2342948, author={Adlian Jefiza and Muhammad Arif Fardiansyah and Ika Karlina Laila Nur Suciningtyas and Diono Diono and Indra Daulay and Dodi Prima Resda and Fadli Firdaus and Jhon Hericson Purba and Agus Kurniawan}, title={Motion Recognition Uses Accelerometer and Gyroscope Sensors Using Learning Vector Quantization Method}, proceedings={Proceedings of the 6th International Conference on Applied Engineering, ICAE 2023, 7 November 2023, Batam, Riau islands, Indonesia}, publisher={EAI}, proceedings_a={ICAE}, year={2024}, month={1}, keywords={learning vector quantization accelerometer gyroscope arduino nano at mega 328}, doi={10.4108/eai.7-11-2023.2342948} }- Adlian Jefiza
Muhammad Arif Fardiansyah
Ika Karlina Laila Nur Suciningtyas
Diono Diono
Indra Daulay
Dodi Prima Resda
Fadli Firdaus
Jhon Hericson Purba
Agus Kurniawan
Year: 2024
Motion Recognition Uses Accelerometer and Gyroscope Sensors Using Learning Vector Quantization Method
ICAE
EAI
DOI: 10.4108/eai.7-11-2023.2342948
Abstract
Determining the correct position and orientation in the technical system has an important role. This determination can be made by combining the use of the Accelerometer sensor with the Gyroscope sensor. This position determination will be carried out to determine the form of movement, namely standing, sitting, lying down, and prayer movements in the form of bowing, prostration, and sitting between two prostrations using the orientation of the x, y, and z axes. This determination uses an artificial neural networks method, namely the Learning Vector Quantization (LVQ) method. In order to retrieve data, a device designed consisting of accelerometer and gyroscope sensors that have been installed on the MPU 6050 sensor with a control system using the Arduino Nano AT Mega 328 microcontroller. Data obtained in formed of values processed to know the accuracy level which Learning vector quantization showing 100% for data training and 72% for data testing.