Related Experiment Video
Updated: May 2, 2026
06:49
Evaluation of a Smartphone-based Human Activity Recognition System in a Daily Living Environment
Published on: December 11, 2015
8.3K
Detection of lying down, sitting, standing, and stepping using two activPAL monitors
David R Bassett1, Dinesh John, Scott A Conger
11Department of Kinesiology, Recreation, and Sport Studies, University of Tennessee, Knoxville, TN; and 2Department of Health Sciences, Northeastern University, Boston, MA.
Medicine and Science in Sports and Exercise
|March 7, 2014
Summary
Adding a second activPAL monitor to the torso accurately distinguishes between sitting and lying down. This dual-monitor approach enhances activity classification for physical activity research.
Area of Science:
- Biomedical Engineering
- Kinesiology
- Physical Activity Monitoring
Background:
- The activPAL (thigh-worn accelerometer) classifies activities but cannot differentiate sitting from lying down.
- Accurate posture detection is crucial for understanding sedentary behavior and physical activity.
Purpose of the Study:
- To investigate if a second activPAL monitor on the torso can differentiate seated from lying postures.
- To improve the precision of activity monitors in posture classification.
Main Methods:
- Fifteen healthy adults wore two synchronized activPAL monitors (thigh and torso).
- Participants performed activities including lying (supine, prone, side), sitting, standing, and ambulation.
- Spatial orientation data from both monitors were analyzed to determine posture and activity.
Main Results:
- The dual-monitor system accurately classified lying down (100%) and sitting (100%).
- Standing/light activity was classified with 90.8% accuracy, and stepping with 100% accuracy.
- Four distinct behaviors were reliably identified.
Conclusions:
- Using two activPAL devices significantly enhances the ability to distinguish between sitting and lying postures.
- This method provides more detailed postural allocation data compared to a single thigh-worn device.
- Improved posture classification aids in more accurate physical activity assessment.
