Validation of Thigh Angle Estimation Using Inertial Measurement Unit Data against Optical Motion Capture Systems

Area of Science:

• Biomedical Engineering
• Robotics
• Human-Computer Interaction

Background:

• Inertial navigation systems commonly use inertial measurement units (IMUs) for human body orientation estimation.
• Existing orientation estimation algorithms are often computationally intensive, posing challenges for real-time embedded systems.
• Accurate orientation tracking is crucial for assistive technologies, such as navigation aids for the visually impaired.

Purpose of the Study:

• To develop and evaluate a computationally inexpensive orientation estimation algorithm (GIOF) for real-time applications.
• To estimate the forward and backward swing angle of the thigh for a vision-impaired navigation aid.
• To reduce computational complexity while maintaining high accuracy in orientation estimation.

Main Methods:

• The Gyro Integration-Based Orientation Filter (GIOF) algorithm fuses accelerometer and gyroscope data for single-dimension orientation estimation.
• GIOF corrects orientation using accelerometer readings when detecting gravity and integrates gyroscope readings otherwise, mitigating drift.
• The algorithm was validated against the Vicon Optical Motion Capture System.

Main Results:

• GIOF achieved a mean correlation of 99.58% with Vicon measurements across 374 walking trials.
• The Root Mean Square Error (RMSE) for thigh angle estimation was 1.8477°.
• GIOF demonstrated approximately half the computation time of the Complementary Filter on an 8-bit microcontroller.

Conclusions:

• GIOF provides an accurate and computationally efficient method for estimating thigh angle using IMUs.
• The algorithm is suitable for real-time embedded systems with limited resources, particularly for vision-impaired navigation aids.
• GIOF's architecture allows for straightforward extension to 2D orientation estimation (pitch and roll).