A real-time collision detection between virtual and real objects based on three-dimensional tracking of hand

In the working environment of argument reality, users interact with virtual objects by hands. Making real-time collision detection between virtual and real objects is very essential to enhance the immersion and authenticity of the environment. In this paper, the key points of hand are marked by color block. We use a pair of common web cameras to collect images and track the points. Then the stereo vision calibration algorithm is achieved. We get hand positioning by calculating the disparity of key points between two images and restoring the three-dimensional position. At the same time, the virtual object models are divided to a binary tree and pretreated by the OBB bounding box. At last, the accurate collision detection is achieved by calculating the relative position between real hand and virtual objects real-time.