CSR:EAGER: A Wearable Body Motion Sensing Platform Using Conductive Stretchable Fabric
This project aims to develop a wearable platform that enables the generation of a three dimensional model of the human skeleton and tracking its motion. There exists techniques that use depth cameras and special image processing software to model motion of the human skeleton, and Kinect is a well-known example. However, the techniques using fixed cameras need to deploy camera devices at fixed locations in order to sense human motions. While this is ok for indoor fixed scenarios, it does not work for outdoor ubiquitous scenarios. Instead, the wearable platform from this project works for both scenarios.
This wearable platform is designed using conductive stretchable fabric that is comfortable to wear. This project exploits the fabric’s resistance change for inferring the bend angle of a body joint that the fabric covers. Joints’ angles are then fused with body skeleton to estimate 3D body motions. The main intellectual merits include: a novel wearable platform that enables ubiquitous body motion sensing, novel user-dependent angle detection sensors with nonlinear mapping models from resistance measurements to joint bend angles, and a novel 3D body motion estimation method that fuses independent joint angles and optimizes estimation results with body kinematics constraints.