Performance Comparison of Xsens DOT and OpenPose in Estimating Posture in Running

In the field of sports, there are many techniques for measuring running posture while the IMU-based Xsens DOT and the computer vision-based OpenPose are commonly used. These techniques have different advantages and limitations. Of these, their validity and accuracy are of particular importance. In this research project, we had 6 participants wear Xsens DOT sensors and run on a treadmill under the observation of a camera. We collected data from the sensors, processed it, and obtained their cadence and vertical oscillation during running. Subsequently, we compared these data. Based on this, we also discussed the use scenarios of these systems to help choose system when designing a running feedback system.