Ancient Chinese-style architecture has received increased attention during the last century as a segment of cultural heritage and is of great significance, specifically in regard to the process of digitizing and modeling these buildings to preserve and protect this heritage. Because the roof form reflects the age of the structure, the structural character and the historical culture of the ancient building, constructing a refined model for the roof is a primary aspect of the 3D modeling procedure. To avoid cumbersome traditional modeling approaches that use geometry units, such as points, lines and triangles, a flexible semantic method is proposed in this study to improve modeling efficiency and reduce the professional requirements. In this method, a two-level semantic decomposition of the roof is presented according to the characteristics of ancient Chinese-style architecture. The structural level reveals the basic components that determine its structural shape, and the decorative level refers to the attached components that influence the exterior appearance. The assembly validity of the decomposed elements and the combined diversity of the integrated entities are ensured by topological constraints and derived transformations of the semantic components. This proposed method was implemented by utilizing CityGML (City Geography Markup Language) via the ADE (Application Domain Extension) mechanism and was tested by modeling the principal buildings included in the Palace Museum.

The segmentation of urban scene mobile laser scanning (MLS) data into meaningful street objects is a great challenge due to the scene complexity of street environments, especially in the vicinity of street objects such as poles and trees. This paper proposes a three-stage method for the segmentation of urban MLS data at the object level. The original unorganized point cloud is first voxelized, and all information needed is stored in the voxels. These voxels are then classified as ground and non-ground voxels. In the second stage, the whole scene is segmented into clusters by applying a density-based clustering method based on two key parameters: local density and minimum distance. In the third stage, a merging step and a re-assignment processing step are applied to address the over-segmentation problem and noise points, respectively. We tested the effectiveness of the proposed methods on two urban MLS datasets. The overall accuracies of the segmentation results for the two test sites are 98.3% and 97%, thereby validating the effectiveness of the proposed method.