Vision-Based Dynamic Tracking of Motion Trajectories of Human Fingertips

摘要

Dynamic tracking of motion trajectories of human fingertips is a very challenging job considering the requirement of accuracy as well as speed. A binocular vision is adopted to dynamically measure the fingertip positions of the human hand. Based on Kalman filter, combined with the historical motion data of human hand, a dynamic tracking model is presented which can fast and accurately track the fingertip positions. The experimental result shows that when human fingers move in a natural speed, the dynamic fingertip positions can be tracked successfully.