Research Of Improving LOD Algorithm In Virtual Reality Based On JAVA3D

Three-dimension virtual reality technology is the cross research field among many subjects. It's one of the most famous research spots nowadays. Because of its advantages such as intuition and interaction, three-dimension virtual reality is applied in modeling and emulating, scientific caculation, visual designing and planning, distance education and distance control, medical science, art and amusement and so on. It has been having more and more unreplacable function in development in science and technology and national economic construction.This subject, the modeling and application of three-dimension virtual reality, is the sub task of "the study of key technology of digital information obtaining in the digital city ", which is the picture information processing and intellectual control Ministry of Education key laboratory open research foundation subject. Major work and content of this paper are as follows:1 , Compared with the popular developing tools nowadays such as VRML, this subject use Java3D to be the developing tool, which bases on Java and is a new three-dimension image tool kit. Java3D inherits Java so many advantages such as OOP and fits for network application, which makes the graph scene made by Java3D has compeletly functions as well as vividly scene. After introduces Java3D and other relative things, the article introduces the whole process of modeling a integrate graph scene from the level of arranging classes and data, through which we confirm that Java3D has the advantage to be applied in the virtual reality technology and has the potential to be applied in the Internet in the future. This also proves that it is novel and correct for us to choose Java3D to be the developing tool.2, At the same time, the important work of the article is to analyse the conventional LOD(Level Of Detail) algorithm and point out its flaws, then put forward the improving method, insteading of the conventional algorithm of modeling many LOD models to render different details, the improving algorithm try to conbine these different details to the integrate model and render them by different stimules,this improving can help us to achieve the goal of reducing the redundant data made by many models while hold the conventional thinking of rendering with different levels, hoping not only the efficiency and effect can reach the level as the conventional algorithm can, but also compress the data which can let it more fit for applying in the Internet. It widens the area in which the virtual reality may use.3 , At last, based on comparing the improving LOD method and the conventional LOD algorithm used in the virtual reality modeling, the article makes the conclusion that the improving method has more advantages than the conventional algorithm, the improving LOD algorithm can reality achieve the goal to reduce the code and data to be transport on the Internet. Then the article points out forward work of this subject and application foreground.