With the concept of the digital city development, and the increasing requirement of virtual reality application in tourist spots and community towns, the traditional ancient building modeling becomes a major research focus in the area of computer-aided design and virtual reality. In view of the demand for current large-scale ancient Chinese building scene automatic modeling, this dissertation researched the automatical modeling of Huipai architectures.On the basis of the typical automatical city modeling methods, a new way of Huipai architectures rapid modeling using module dissecting and BFS graph traverse algorithm was introduced in this dissertation. First of all, Huipai architectures' modules were parameterized into nodes of topological graph, making model's shapes and modules' appearance can be controlled by those specific parameters. Then, the combination rules of Huipai architecture were summarized according to real Huipai houses characteristic. These rules were used for subsequent module combination detecting, making modeling more reasonable. Furthermore, Huipai architecture models were constructed automatically by traversing these nodes using BFS algorithm. During the traversing process, the pre-created combination rules were used to detect combination errors, so that the modeling of Huipai architectures were applied to the actual Huipai architectures. Finally, a new method of automatical topology generation was discussed and studied based on Bayesian Network theoretically, and this method can translate architecture modules into graph topology automatically, which is helpful to construct Huipai architectures more quickly and automatically!... |