**1. Introduction**

Traditional Chinese sport has been a compulsory component of Physical Education (PE) in universities in China since 2002 [1]. Although there are various traditional Chinese sports to choose from, 76.7% of universities taught martial arts in their PE curriculum [2]. In 2016, the Communist Party of China and the Chinese governmen<sup>t</sup> adopted the 'Healthy China 2030' national health plan [3]. In this plan, Baduanjin was identified as a traditional Chinese sport that was promoted and supported by the government. This resulted in increased Baduanjin teaching and research in universities throughout the country [4].

Although universities in China must incorporate traditional Chinese sports into their PE curriculum, there have been problems with its implementation. These include a high student-teacher ratio, uninteresting forms of teaching-learning resources, and an incomplete assessment system. These three problems adversely a ffected the requirements for teaching quality set by the People's Republic of China Ministry of Education [5,6]. Although the high student–teacher ratio has been a problem since 2005, it has ye<sup>t</sup> to be resolved [7,8]. Teachers are not able to provide individual guidance to each student because of the large number of students in the class. As a result, teachers cannot correct all the students' mistakes, and students are not aware of their incorrect movements [9].

In recent years, motion capture (Mocap) has been widely applied in fields such as clinical and sports biomechanics to distinguish between di fferent types of motions or analyze di fferences between motions [10,11]. Studies have also applied Mocap in PE for adaptive motion analysis to evaluate the motion quality of learners and feedback the information to assist them in detecting and correcting their inaccurate motions. In the study by Koji Yamada et al. [12], a system based on Mocap was developed for Frisbee learners. Researchers used the Kinect device to obtain 3D motion data of learners during exercise, detect their pre-motion/motion/post-motion, and display the feedback information to improve their motions. The results showed that the system developed by the researchers can e ffectively improve the motions of learners [12]. Chen et al. [13] applied Kinect in Taichi courses in universities. The 3D data of motions from novice students captured by Kinect were compared with an expert in order to evaluate the quality of motions and students were informed of their results. The research showed that the motion evaluation system on Kinect developed by the researchers accelerated learning by novice Taichi students. More recently, Amani Elaoud et al. [14] used Kinect V2 to obtain red, green, blue and depth (RGB-D) data of motion. They used these data to compare the di fferences between novice students and experts on the central angles of points that a ffect throwing performance in handball. These experiments show how researchers have used various categories of Mocap.

Based on di fferent technical characteristics, the application of Mocap in PE can be divided into four categories: optoelectronic system (OMS) [15], electromagnetic system (EMS), image processing systems (IMS), and inertial sensor measurement system (IMU) [16]. In these four Mocap categories, OMS is the most accurate and is considered to be the gold standard in motion capture [16,17]. However, OMS requires a large number of high-precision and high-speed cameras that will inevitably result in issues related to cost, coordination, and manual use [18]. Moreover, OMS cannot capture the movement of objects when the marker is obscured [19]. These deficiencies have limited the practical application of OMS in PE. The advantage of EMS over OMS is that it can measure motion data of a specific point of the body regardless of visual shielding [20]. However, EMS is susceptible to interference from the electromagnetic environment which distorts measurement data [21]. Also, EMS has to be kept within a certain distance from the base station, which limits the use range [22]. IMS has better accuracy compared to EMS and an improved range compared to OMS [16]. Most studies have used low-cost IMS (such as the Kinect device) to capture motion for analyzing motion in PE. However, there are some disadvantages in low-cost IMS, namely low-accuracy, insu fficient environment adaptability, and limited range of motion because the Kinect sensor has a small field of vision [16]. High-performance IMS does not have these shortcomings. Generally, high-performance IMS has favorable accuracy and a good measurement range. However, high-performance IMS requires expensive high quality and/or high-speed cameras which has limited its application [16].

Based on the disadvantages of low-cost IMS in its application in PE, applying IMU (a motion capture consisting of an accelerometer, gyroscope, and a magnetometer) in PE may mitigate these application problems [23]. In recent years, the development of technology has reduced the cost of IMU, making it possible to be used in PE. The validity of assessing motion accuracy of IMUs has been confirmed. Poitras et al. [24] confirmed the criterion validity of a commercial IMU system (MVN Awinda system, Xsens) by comparing it to a gold standard optoelectronic system (Vicon). Compared to low-cost IMS, IMU has certain advantages in environmental adaptability and a su fficient range of motion. The IMU does not require any base station to work, which means it is the most mobile of the available motion capture systems [16]. Moreover, IMU can measure high-speed movements and is non-invasive for the user, making it an attractive application for PE [16,25].

However, there are a few issues capturing motions using IMU. First, the IMU sensors are sensitive to metal objects nearby which distort the measurement data [16]. Therefore, participants should wear fewer metal objects when capturing motion using IMU. Fortunately, in traditional Chinese martial arts such as Baduanjin, exercisers should, in principle, wear traditional Chinese costumes without any metal, which minimizes the impact of metals on IMU sensors. Second, a common IMU system, Perception Neuron 2.0, was used in our research which uses data cables to connect all sensors with a transmitter. Although users cannot wear this wired IMU system on their own, it does not a ffect the

accuracy of the data. Also, the latest IMU overcomes this problem that users can't wear it on their own by having each sensor transmit the data to the external receiving terminal directly [26].

Therefore, we propose applying IMU in Baduanjin by developing a system that assesses and records the quality of motions to assist teachers and students in determining inaccurate motions. Using IMU, students can learn Baduanjin independently after class and teachers can evaluate students' progress, which is useful for formative assessment. That may alleviate current problems faced in PE classes in Chinese universities. For this purpose, we explored the feasibility of using an IMU to distinguish the di fference in motion accuracy of Baduanjin between novice and senior students.

#### **2. Materials and Methods**
