Ancient Architecture Digital And Key Technologies Of The Three-dimensional Modeling Study

Compared with the achievement of advanced construction technology in ancient China, information technology of our country in the ancient buildings are still relatively backward. So it is needed to conduct some researches on protection and development on ancient buildings using modern information technology. In this paper, based on the National Natural Science Foundation of "three-dimensional model of cultural ontology semantic annotation and retrieval technology" (ID:60873094) and National 863 Project "three-dimensional model on intelligent processing and retrieval platform" (ID:2008AA01Z301), some key technologies like engineering structure 3D model, point cloud data 3D model and 2D image reconstruction are studied and some achievements have been made. Researches in this paper are as follows:1. According to the study on the ancient architectural form of integration, an universal component library of ancient buildings are established with a single inch (tou-k'ou modules, the eleventh grade) as a unit module. Further more, the visual application interface of component library is developed.3D model of ancient buildings can be fast reconstructed based on surface model, solid model and integrated model.2. Proposed a parametric 3D model reconstruction method of ancient buildings. This method, which is based on the ancient architectural drawings, studied the semantic description of solid elements, size marking and latent pattern information to obtain relative parameters. These parameters can call the 3D components from the library, so the goal of fast reconstruction of 3D model of ancient buildings has been achieved. The proposed method established the engineering level of the precise vector 3D model that contains the image, intuitive, accurate digital information, and provided a new technology for the reconstruction and repair on ancient buildings. To verify this method, three-dimensional model of the hip roof is reconstructed and showed good result.3. Proposed a point cloud data reduction algorithm based on the clustering plane feature. This method, which doesn't need surface reconstruction on scanning object, analyzed the hierarchical clustering data on condition of maintaining the precision of building height, retained the feature points which meet the condition and deleted the other points to achieve the data reduction of building point cloud. The instance showed that this method can achieve 70%-90% reduction ratio according to different shape characteristics, and greatly improved the efficiency of the 3D reconstruction of large-scale ancient buildings.4. Proposed 3D coordinate reconstruction algorithm based on basic elements'projection and vanishing point calibration principle. According to the characteristic of ancient buildings, the algorithm was further improved by giving voxel reconstruction calibration. Firstly, the 3D coordinates of basic feature points were obtained by projection inverse algorithm, then the projection matrix were conducted and the camera were calibrated according to obtained 3D coordinates, so the reconstruction from 2D images to 3D model was successively made. The modeling of white cabinet of Small Wild Goose Pagoda ancient buildings showed that this algorithm was effective.