Real-Time Foot Height Estimation and Activity Classification Using a Foot-Mounted IMU Implemented on a Smartphone.

Area of Science:

• Biomechanics
• Wearable Technology
• Human Movement Analysis

Background:

• Wearable sensors enable continuous gait assessment, crucial for evaluating fall risk.
• Inadequate foot clearance during locomotion is a primary cause of tripping and falls.
• Existing methods for foot height measurement using inertial measurement units (IMUs) suffer from cumulative drift errors and lack real-time activity classification.

Purpose of the Study:

• To develop a real-time, single-IMU system for accurate foot height trajectory reconstruction.
• To simultaneously classify five locomotion activities using smartphone-based IMU data.
• To provide a practical foundation for real-time gait intervention and fall prevention.

Main Methods:

• A real-time system utilizing a single IMU integrated into a smartphone was developed.
• Kinematic constraint-based approaches were employed for foot height trajectory reconstruction.
• A convolutional neural network was used for simultaneous classification of five locomotion activities.

Main Results:

• The system achieved accurate foot height measurement with minimal cumulative errors (<1.1 cm) across various terrains.
• Mean absolute error for height estimation was 0.42%, with level walking maintaining ground reference.
• The system demonstrated 96.08% overall classification accuracy for locomotion activities with minimal latency.

Conclusions:

• The developed real-time system offers a practical solution for continuous foot height monitoring and gait analysis.
• This technology has significant potential for applications in fall prevention and personalized gait interventions.
• Simultaneous activity classification enhances the utility of wearable sensors for comprehensive gait assessment.