A Visual Glove Based on Conformal Geometric Algebra

In this paper, a visual glove based on Conformal Geometric Algebra is proposed. To estimate the motion of hand in an intuitive way, Conformal Geometric Algebra is introduced in that it can describe relationship of objects in 3D Euclidean space geometrically. For measuring human hand joints angles with glove like device, the physiological structure of human hand is analyzed and an inverse kinematic model is constructed for obtaining the joint angles of fingers. In the Conformal Geometric Algebra framework, a visual glove without sensors and signal process parts on itself is designed. The features on glove surface are extracted in a binocular computer vision system, and the self-calibration algorithm for the glove is implemented. With the calibrated glove, all the geometric objects for inverse kinematic computing by Conformal Geometric Algebra method are reconstructed in 3D space. With the support of this visual glove, the motion of hand model is demonstrated in the virtual environment.