A Novel Measurement Procedure for Error Correction in Single Camera Gait Analysis

Abstract

Gait analysis (GAn) is crucial for assessing human biomechanics, aiding rehabilitation, and informing health diagnostics. Traditional optoelectronic systems represent the gold standard for determining joint kinematics, but they present challenges due to high costs and technical demands in offering precise 3D marker-based GAn. Inertial measurement units (IMUs) provide a portable, cost-effective alternative but face data volume, setup time, and calibration challenges. Thanks to recent advancements in computer vision, markerless video-based solutions have become increasingly popular as they have the potential to overcome the limitations of marker-based systems and IMUs. In this study, we developed and validated a novel sensorless method for estimating knee flexion and ankle dorsiflexion angles using a single-camera setup integrated with the OpenPose (OP) open-source library. Thirty-one healthy subjects (20 females, 11 males) were recorded walking on a sensorised treadmill with lower-body IMUs at a natural pace, using a synchronised webcam. Results showed that the markerless software demonstrated significant accuracy in estimating the angles of knee flexion and ankle dorsiflexion compared to IMU-based measurements. By applying a universal compensation algorithm, we demonstrated that the method is subject-independent, ensuring its robustness and generalisability across different populations. Results showed that the markerless software demonstrated significant accuracy in estimating the angles of knee flexion and ankle dorsiflexion compared to IMU-based measurements. After compensation, the method achieved a compatibility of over 93% for knee joint angles and significant improvements for ankle dorsiflexion, indicating its potential to complement or substitute traditional IMU systems in biomechanical assessments. This validation suggests that our cost-effective, markerless approach can complement or substitute traditional IMUs systems in biomechanical assessments.

Keywords

• Biomechanics
• gait analysis
• inertial measurement units
• joint angles estimation
• markerless motion capture