JTIM : Jurnal Teknologi Informasi dan Multimedia
Vol 1 No 4 (2020): February

Wearable System untuk Mendeteksi Aktivitas Fisik Menggunakan Nilai Ambang Batas Akselerasi

Made Liandana (ITB STIKOM Bali)



Article Info

Publish Date
18 Feb 2020

Abstract

Wearable devices can be used to detect activities. The size of the wearable device is relativelysmall and easy to wear on clothing so it does not complicate the movement of its users. In thisstudy wearable devices or wearable systems were developed using a three-axis accelerometersensor. The detected activities are walking, sitting and lying down. The technique used todetect activity is to use a threshold value. For walking activities use the y axis accelerationwhose value has been reduced from the influence of gravity. The threshold value for thecurrent activity consists of two, namely: the lower threshold value and the upper thresholdvalue, the lower threshold value is between -1 to -2 and for the upper threshold value rangesfrom 1 to 2. The sitting and lying activity uses the value acceleration on the x, y, and z axes.For sitting activity the x-axis threshold value is between -0.5 to 0.5, the y-axis value is between9 and 10, and the z-axis value is -0.5 to 0.5. The threshold value of lying activities on the xaxis is between -0.5 and 0.5, the y-axis is between -0.5 and 0.5, while the z-axis is between 9and 10. The test results on wearable devices used to detect walking, sitting and lying activitiesshow, the percentage of results of running activities by 60%, for testing sitting activity by 65,while for lying down activity by 60%, each test is carried out 20 times.

Copyrights © 2020






Journal Info

Abbrev

jtim

Publisher

Subject

Computer Science & IT

Description

Cakupan dan ruang lingkup JTIM terdiri dari Databases System, Data Mining/Web Mining, Datawarehouse, Artificial Integelence, Business Integelence, Cloud & Grid Computing, Decision Support System, Human Computer & Interaction, Mobile Computing & Application, E-System, Machine Learning, Deep Learning, ...