A Human Gait Tracking System Using Dual Foot-Mounted IMU and Multiple 2D LiDARs

Area of Science:

• Robotics and Human-Computer Interaction
• Sensor Fusion and Navigation

Background:

• Inertial Measurement Units (IMUs) suffer from drift errors in gait tracking.
• 2D LiDAR systems have limited tracking range.
• Combining IMUs and LiDARs can overcome individual sensor limitations.

Purpose of the Study:

• To propose a robust human gait tracking system by fusing dual foot-mounted IMUs and multiple 2D LiDARs.
• To mitigate IMU drift errors and extend LiDAR tracking range.
• To enable accurate gait analysis without extensive calibration.

Main Methods:

• Utilizing two 2D LiDARs as anchors to correct IMU-based inertial navigation.
• Employing a Kalman filter and smoother algorithm to fuse sensor data.
• Implementing a novel method for calibrating spatially separated LiDAR units.
• Using initial LiDAR scans for IMU heading and position estimation.
• Incorporating stride heading and step width constraints for enhanced accuracy.

Main Results:

• The proposed system accurately tracks human gait by correcting IMU drift with LiDAR measurements.
• Initial heading and position estimation for IMUs achieved without prior calibration.
• Successful calibration of widely separated LiDAR sensors.
• Stride length estimation error below 3 cm (RMSE) compared to optical systems in straight walking.
• Validation of the system's performance in both straight and rectangular path walking scenarios.

Conclusions:

• The dual IMU and multiple 2D LiDAR system offers a robust solution for human gait tracking.
• Sensor fusion effectively overcomes the limitations of individual IMU and LiDAR systems.
• The method provides accurate and reliable gait parameter estimation for various walking paths.