Vision-based Swing Trajectory Estimation using RGBD Camera

Human movement analysis is a significant topic for understanding and improving human activities. In this paper we propose a vision-based swing trajectory detection scheme. The goal of the proposed idea is to trace the trajectory of swung object using an RGBD camera. The coordinates of the tip of grasping object are estimated from hand joints in pose estimation. The feasibility of the proposed scheme was demonstrated via experimental results of cleaning activities using a mop. It will contribute for assessing users' progress by giving further training tips for swing motions.