Game Engine Based On The Three-dimensional Positioning Of The Realization Of The Interactive System

With the rapid development of computer technology , the requirements of human-computer interaction increasingly high. New kinds of interaction could get rid of the space limitation of the keyboard and mouse so as to achieve the exchange of information and control exchange through a natural way. This requires that the computer must have the intelligent which could able to imitate people's vision, hearing and so on. This article focused on space location, by using the theory of binocular vision and identifying and tracking the characteristics of movements so as to determine the information about the location of objects. And then, judge the person's instructions through the differences between the information about the location.There are a lot of ways about Natural interaction technology. At present, a large number of researches exist in this area, such as gesture recognition interaction and so on. I focus on positioning the three-dimensional system about real-time interactive based on binocular vision in the game engine, proving a real-time three-dimensional interactive way for the games. The system uses cheap cameras as the platform for this paper. In addition, it supplies platform calibration, real-time three-dimensional space which determines the location, three- dimensional attitude determination and so on. Because of implement in the game engine, the efficiency should be much more important, the complex computing algorithm is undesirable. This system uses a specific thing as the target which captures images of the movement of object by the two real-time cameras which were corrected. To begin with, I simply process images, isolating the object from the regional information, in addition, on one side, using the frames to track the trend of the movement from the object, on the other hand, recovering the marked region extracted by two cameras to the original three-dimensional space position and posture of the object through space geometric method. A last, complete both sides of the information to implement the three-dimensional object tracking. This method can be extended to multiple-vision, so as to expand the movement of the range of marked object, but the first important thing is to solve the problem about the input interface of cameras. In this system, I use a host of the technologies about digital image processing and project calculation. Finally, this system is used as a plug-in to the AI editor of the digital content creation platform, so as to achieve the interactive simulation of the intelligent behavior to the role's reaction.We design the software of the system by using a standard design procedure, which optimize the architecture of the system and the algorithm.First of all, this thesis analysis and summaries all the related technologies, then, it gives the details about the architecture design and treatment flow of the system, moreover, it discuss the theory of the algorithm which was used by this paper. At last, it gives the result of the experiment.