Research And Implementation Of Key Technologies For Vision Immersion In Interactive Control Of Virtual Reality

Distributed virtual reality technology, which is one of the most powerful man-machine interactive technologies, has attracted great research interests in the field of information. Visual immersion is the main design purpose for a distributed virtual reality system. Visual immersion of the traditional virtual reality system mainly focused on the static scenes, such as fine level, lighting effect and texture mapping of models. With the development of distributed virtual reality system, the user groups keep growing, the virtual environments are more and more complex, and the interaction between virtual environment and users become in-depth. The dynamic problems, which affect the visual immersion, have become more and more evident. Therefore, improving the visual immersion of virtual reality system, particularly the visual immersion of interactive control with the users in large scale and complex scenes, has become an important challenge in distributed virtual reality system.In this paper, the key technical issues affecting the visual immersion of the interactive control in distributed virtual reality system have been studied, including vision in real-time and consistency in distributed scenes, the accuracy and real-time of interactive control between user and virtual scene, and so on. The key technologies of improving the visual immersion of intelligent interactive control in distributed virtual reality system have been designed and implemented. Immersion effect both in vision and sense have been remarkably enhanced through the feedback control studies of 6-DOF parallel robot platform.The main work and innovations of this paper are listed as follows:(1) A new hybrid P2P architecture framework for the distributed virtual reality system has been designed and implemented. The whole system were logically divided into three parts: the centralized P2P network structure, the centralized P2P grouping and the structured distributed P2P network team.Three types of nodes, including the network service platform, control platform and rendering node, have been achieved, which improved the transmission speed of information, and reduced the delay of visual information;(2) Rendering strategy of different types of nodes based on P2P network structure has been designed and implemented. For the same team of nodes, render strategies of main-follow and cache memory were proposed to meet the visual consistency of the same sub-block scene. For the same group of nodes, the consistent control and prediction algorithm were proposed to meet dynamic synergy between the multiple users consistently. So the consistency of the visual effects was confirmed;(3) The collision detection algorithm of honeycomb-like space decomposition based on ray casting has been designed and implemented. The collision detection has been changed into the problem of comparing the distance between local sub-polyhedron cross section and the value of opponent in the next frame with the set value of opponent. It completed the fine detections between the polyhedrons at random which not only increased the speed of collision detection but also enhanced the visual impact;(4) The collision detection algorithm based on the basic point has been designed and implemented. Though calculating the intersection of scanning lines and disparting space polygon by "go-between" approach, the algorithm has changed the collision detection between models into the detection between pairs of basic geometric characteristics, which was more suitable for the interactive control system of fast, real-time, and precision. This algorithm improved the truth of the visual effect;(5) The dynamic model of motion platform of a new type of 6-DOF parallel robot has been designed and implemented. For the structural features of the new type of 6-DOF, the inverse kenematics of the system has been solved. The motion equations of platform were obtained by the Lagrangian. Dynamic equations of 6-DOF parallel mechanism were also established by Lagrange's equation. Then simple processing was developed in accordance with the application of the model in the actual control. This model completed the feedback control of the virtual reality, and it will further enhance the user's immersion in vision and sense.