Research on Thangka Image Scene Switching Based on VR

With the development of computer simulation technology and computer graphics, virtual reality (VR) has become the hotspot and difficulty in the current world research. This paper embarks from the actual and presents a Thangka image browsing research based on VR. The second order gradient enhancement of Sobel operator algorithm, maximum entropy segmentation algorithm, the most gray value segmentation algorithm and point to linear symmetry method are used to realize the VR-based Thangka image scene switching. Experimental results show that the processing time obtained through Leap Motion is 20-30 ms/frame, and the accuracy of rigid body region detection is more than 70%. It can basically meet the requirements of real-time and accurate handoff of Thangka image scene.