GHand: A GPU Algorithm for Realtime Hand Pose Estimation Using Depth Camera

Abstract

We present GHand, a GPU algorithm for markerless hand pose estimation from a single depth image obtained from a commodity depth camera. Our method uses a dual random forest approach: the first forest estimates position and orientation of hand in 3D, while the second forest determines the joint angles of the kinematic chain of our hand model. GHand runs entirely on GPU, at a speed of 64 FPS with an average 3D joint position error of 20mm. It can detect complex poses with interlocked and occluded fingers and hidden fingertips. It requires no calibration before use, no retraining for differing hand sizes, can be used in top or front mounted setup and with moving camera.