A Comparison of the Performances of Video-Based and IMU Sensor-Based Motion Capture Systems on Joint Angles

Abstract

The optoelectronic system gives the most sensitive results among motion capture systems; however, it has limited use due to its disadvantages such as high cost, time-consuming calibration process, and space limitation. On the other hand, the inertial measurement unit (IMU) sensors are lightweight, portable, low-cost, and easy-to-implement systems consisting of small components. Rokoko Smartsuit Pro is a wearable suit consisting of an array of IMU sensors often used in the movie industry. In addition, it is considered that it can be helpful in sports biomechanics with its ease of application because it has the advantages of IMU sensors. However, its applications in sports biomechanics are not entirely accepted due to uncertainty in the reliability of motion capture. This study aims to its applicability and reliability in sports biomechanics by comparing the system with Kinovea, which is proven, open-source, and easy to use. Motion data were collected from five subjects using the Rokoko and smartphone camera simultaneously for joint angles of the lower limbs. It was recorded flexion/extension movements of the hip and the knee, plantar flexion/dorsiflexion movements of the ankle. The results were statistically analyzed by comparing them with the Bland-Altman using SPSS software. In addition, the root-mean-square error (RMSE), Pearson correlation, and independent t-test were calculated. Accordingly, the RMSE was a maximum of <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$3.51 {}^{\mathrm{o}}$</tex> angulation, the Pearson correlation was obtained as 0.96, and no significant difference was found between the systems (p<0.06). As a result, it can be utilized for sports biomechanics because the proposed system is an economic rather than an optoelectronic system and provides highly accurate and reliable results.