Related Experiment Video
Updated: Jan 11, 2026
07:25
An Objective and Child-friendly Assessment of Arm Function by Using a 3-D Sensor
Published on: February 12, 2018
7.2K
Non-Contact Fatigue Estimation in Healthy Individuals Using Azure Kinect: Contribution of Multiple Kinematic Features
1National Institute of Technology, Tsuyama College, 624-1 Numa, Tsuyama 708-8509, Japan.
Sensors (Basel, Switzerland)
|November 13, 2025
Summary
This study shows that non-contact camera-based analysis of body movement can accurately estimate exercise-induced fatigue. This approach helps optimize training and prevent injuries by monitoring fatigue levels during physical activity.
Area of Science:
- Biomechanics
- Sports Science
- Human Movement Analysis
Background:
- Monitoring exercise-induced fatigue is crucial for effective training and injury prevention.
- Current methods for fatigue assessment can be invasive or subjective.
Purpose of the Study:
- To evaluate a non-contact method for estimating perceived fatigue using full-body kinematics.
- To assess the feasibility of using camera-based motion analysis for fatigue monitoring.
Main Methods:
- Ten healthy adults performed reproducible whole-body movements while 3D skeletal data was captured using an Azure Kinect depth camera.
- 24 kinematic features were extracted from the skeletal data.
- A random forest classifier was trained to predict fatigue levels (Low, Medium, High) based on kinematic features, with class imbalance addressed using oversampling.
Main Results:
- The model achieved high performance in discriminating fatigue levels, with an overall accuracy of 86% and a macro ROC AUC of 0.98.
- Feature importance analysis revealed contributions from various kinematic features.
- The non-contact approach demonstrated feasibility for estimating perceived fatigue during simple movements.
Conclusions:
- Camera-based kinematic analysis offers a viable, non-contact method for estimating perceived exercise-induced fatigue.
- This technology has potential applications in optimizing training regimens and preventing sports injuries.
- Future research will focus on expanding the study cohort, diversifying movement tasks, and integrating physiological signals for improved model generalization and interpretability.
