Computer Vision Aided 2D Motion Sequence Test Supervision

摘要

This study presents an object trajectory tracking system that would increase the reliability and accuracy of motion sequence testing for devices such as accelerometers. By attaching an ArUco marker to the device, we are able to perform pose estimation for the device and calculate the coordinates of the device in a coordinate system of our choosing. By tracking the coordinates in consecutive frames, the trajectory is noted. While testing, speed corrections are displayed. We are able to compare the test trajectory to a reference trajectory and determine whether the test device is malfunctioning, or the motion was performed incorrectly. In case of an incorrect trajectory, the exact section that was incorrectly performed is determined, and rectified in subsequent testing. This eliminates uncertainty between the test device malfunction and incorrect device motion performed during the test, thus providing more accurate test results.