Research On The Construction Of Virtual Indoor Environment

As a component of the virtual battlefield environment, it is indispensable to study virtual indoor environment in the simulation of the battlefield environment. This paper build a real-time rendering, immersing and near-real virtual indoor environment, focusing on some key technologies, such as rapid collision detection, collision response, data organization, scene data organization and scheduling as well as the calculation of lighting and shadow in indoor scene.The following contents are what we have done in this dissertation:1. By summarizing some common collision detection algorithms, we present a method to simplify collision intersection with image depth detection in indoor virtual environment. We implement a fast collision detection algorithm with OpenGL-based pick technology. And we also optimize the algorithm combining object space acceleration method. Finally we demonstrate its feasibility by experiment.Combining the radial intersection technology, we implement the terrain following in virtual indoor environment.2. By a research on collision response and its physical models, we find out the key elements that affect the collision response. It provides theoretic accordance for the model data reorganization, and is also the foundation of the application of the indoor environment simulation. And we implement the door's response to viewpoint collision (on and off) and validate the rationality of reorganization.3. Based on the high hidden attribute of the indoor environment, we design a hierarchical structure of buildings, floors and rooms to meet the need of real-time data requirement and present a method of data scheduling based on viewpoint position. At last, we validate its feasibility by experiment.4. We try to explore ways to improve the scene reality considering the characteristics of indoor environment. For example, we improve the light rendering effect with the method of getting the vertexes' normals by the mean value of the faces' that satisfied the constrained condition; and we perfect the scene vision by using shadow volume technology to render shadows in the scene.