Comparative Analysis of Computer Vision and IMU Sensing Systems for Accurate Determination of Gait Parameters

This is the authors' abstract. We don't add key points for this paper.

Gait analysis is a crucial technique in the fields of sports, biomechanics, and rehabilitation. The two primary technologies used in gait analysis are vision systems and inertial measurement units (IMUs). While vision systems are excellent at giving precise and comprehensive gait data, IMUs are well-suited for portable and affordable applications. There are other marker-based systems for gait analysis, but they are very expensive and one cannot access them easily. To completely realize joint angle analysis and biomechanics' potential using marker-less technology, this work make use of a BlazePose machine learning model. IMU devices attach to four separate locations on the patient's body using the MPU6050 to obtain joint angles while the subject is walking. This paper examines the reliability, precision, and applicability of these two technologies in various scenarios. Our findings indicate that while IMUs are more affordable and portable, Vision Systems are also affordable, more accurate, and provide a greater amount of data to enable comprehensive gait analysis.