Related Experiment Video
Updated: May 25, 2026
06:52
An Inertial Measurement Unit Based Method to Estimate Hip and Knee Joint Kinematics in Team Sport Athletes on the Field
Published on: May 26, 2020
Estimation of end point foot clearance points from inertial sensor data
Braveena K Santhiranayagam1, Daniel T H Lai, Rezaul K Begg
1School of Sport and Exercise Science and Institute of Sport Exercise and Active Living, Victoria University, Melbourne, Australia. braveena.santhiranayagam@live.vu.edu.au
Summary
This study accurately estimates foot clearance parameters, like maximum vertical clearance and minimum toe clearance, using inertial sensors. This technique helps assess tripping risks during walking.
Area of Science:
- Biomechanics
- Gait Analysis
- Wearable Sensors
Background:
- Foot clearance during walking is crucial for preventing trips and falls.
- Accurate measurement of foot clearance parameters is essential for gait analysis and fall risk assessment.
- Traditional motion capture systems are complex and limit real-world applications.
Purpose of the Study:
- To develop and validate a technique for estimating key foot clearance parameters using inertial sensor data.
- To assess the accuracy of the proposed method compared to optoelectronic motion capture.
- To investigate the effectiveness of feature selection for improving estimation accuracy and reducing computational load.
Main Methods:
- Inertial sensor data (accelerometers and gyroscopes) were collected from eight subjects walking at four different speeds.
- Fifteen features were extracted from the raw sensor data.
- General Regression Neural Networks (GRNN) and a Leave-One-Subject-Out cross-validation method were employed for parameter estimation and model selection.
- A hill-climbing algorithm was used for feature selection.
Main Results:
- The GRNN model accurately estimated maximum vertical clearance (m x 1) with an average Root Mean Square Error (RMSE) of 5.32 mm and minimum toe clearance (MTC) with an average RMSE of 4.04 mm at maximum walking speed.
- Feature selection improved RMSE by 0.54-21.93% and reduced the number of required input features.
- The proposed method demonstrated high accuracy comparable to motion capture systems.
Conclusions:
- Inertial sensor data, combined with regression models and feature selection, provides an accurate and feasible method for estimating key foot clearance parameters.
- This technique has potential applications in real-world gait monitoring and fall risk assessment.
- The findings support the use of wearable sensors for non-invasive gait analysis.
