A Data-driven 3D Animation System for Tele-Rehabilitation

In this paper, a novel data-driven 3D animation system is developed for stroke patients to use remotely at home environments to recover their motor function. This system consists of four main components, namely human motion tracking, data analysis diagram, data-driven arm animation, and text-driven speech animation. It also contains an expert database and transmission of data files over the Internet. Both patients and therapists can use this system to compare 3D mimic animation with historical case studies of normal and abnormal motions. We have evaluated this system using several simulations: (1) 3D arm rendering in different movement patterns, such as reach, drink, and reach-flexion; (2) data analysis using the diagrams; (3) giving an evaluation and suggestion by the digital human model with a speech animation.