Hand Tracking Guidance by 3D Printing

Research on hand tracking technology in human computer interaction has been actively researched. Computer vision technology based on 2D and 3D cameras is applied for hand tracking. We can improve hand tracking performance by making customized guidance to existing computer vision based hand tracking technology. In this paper, we propose hand tracking system using a guidance produced by 3D printing. Personalized guidance for hand tracking is made using 3D printing. We make the guidance that can be worn between two fingers. The proposed guidance is made into a circular shape using green filaments. A specific color is detected to create a candidate region of the guidance, and a circular shape is found in the candidate region to find the position of the guidance. The center point of the found guidance is tracked through the Kalman filter. In the tracking using the Kalman filter, only motion within the range is tracked by limiting the allowable range of motion. If the tracking fails in the current frame, it memorizes the motion restriction area and resumes tracking when the guidance enters the motion restriction area again. Experimental results show that the proposed system works reliably.