Research On Fast Collision Detection Algorithm Based On Cloud Model And The GPU Cache Technology

The study of collision detection has been long in computer graphics, robot motion planning,3D gaming, artificial computer-aided design, and the field of virtual manufacturing equipment and artificial intelligence. In recent years, with the rising of virtual reality, distributed, parallel, and other technologies, collision detection has been being popular research issues. Now, with the rapid development of computer technology, the geometry creations of the object models are also more and more complex, which increase the computational complexity of collision detection, consume more available resources to the computer, reduce computer running speed, and produce some affection to the execution of other tasks. In the face of the huge computation of the collision detection, many experts and scholars at home and abroad throw themselves into algorithm studies, combing the latest computer technology and their own knowledge, to meet the real-time and accuracy of the algorithm based on a wide range of exploration, and enable them to universal applicability for collision detection of development, pointed out the new research direction for it as well.In an interactive system, real-time interactive is the basic requirements of users system. Therefore, it is required extremely strict for the real-time and accuracy of the collision detection and responses to user instructions timely and accurately. In this essay, the author has studied the related computer technology and then put forward the following two algorithms by combining the basic geometrical element collisions detection algorithm:(1) From the standpoint of reducing computer calculation amount, the calculation model based on cloud fast collision detection algorithm has put forward. Its basic principle makes full use of the advantages of network resources, and centralizes computer network terminal free computing performance to handle the enormous amount of computation during the collision detection. This algorithm without limitations of the number of space objects is suitable for the calculation of large and complex scenes, and better deals with large-scale computation.(2) From the perspective of dealing with large complex scenes, this paper presents a insensitive virtual scene. The basic idea is to use the advantages of GPU graphics processor and gets the conclusions of the collision between the objects by caching technology to analyze images of objects in a scene. For it is not sensitive to the image scene, it is very favorable for the collision detection in large-complex scenes.