Research On Real - Time Sports Simulation Technology With Realism

In recent years, with the development of computer hardware and the decrease of cost,especially the rapid development of GPU and other graphics technology, the computer candisplay more complex content. Therefore, the crowd simulation and its applications hasbeen developed quickly and been used extensively. For instance, it can be applied toemergency evacuation to reduce the loss after the crisis in public places. Since thesimulation by using particle or simple models is unable to meet the needs of people, higherrequirements are put forward for obtaining result of crowd simulation with high degree ofvisual realism. The photo-realistic crowd simulation has become a hot research field.In current crowd simulation systems, the models and motions are basically made byartists manually. Because the way to create them is very difficult, it would take a lot oftime, manpower and financial resources in order to achieve high realistic crowd simulation.With the purpose of cutting cost of man power, some systems use few models and motions,so the characters and the corresponding behaviors are generally lack of variety. Meanwhile,with the development of mobile devices and its applications, the crowd simulation is notonly used in computers. People can easily use related applications in their own mobiledevice. It increases the scope of use, and promotes the development and popularization ofrelated applications.In order to solve the above-mentioned problems, aim for the several major steps incrowd simulation including human modeling, human motion, path planning and thereal-time crowd simulation system. In the last, this paper implements a photo-realisticcrowd simulation system. The main work and innovation of this paper are summarized asfollows:1. Adopt Kinect to obtain a variety of lifelike human models quickly, which canresolve the problems of lack of variety and low creating efficiency in creating modelsmanually. And adopt LOD technology to resolve the low performance problem of massivecrowd simulation using models scanned by Kinect.2. Process the C3D file which is recorded by motion capture device to obtain a largenumber of realistic motions. This method can resolve the time-consuming problem andimprove the effect of motions which are created manually. 3. Propose a file format called extendible path file format to improve the performanceof loading and resolve the problem of loading different path generated by differentalgorithm.4. According to the approach above, relating the projects under research, design thecross-platform real-time crowd simulation system architecture, and implement thecross-platform real-time crowd simulation system based on XNA/MonoGame platform,which can achieve a photo-realistic real-time crowd simulation and resolve the problem ofcrowd simulation in different platforms.