Combination of Region Growing and TIN Edge Segmentation for Extraction of Geometric Features on Building Facades

The 3D reconstruction of building facade can be used for a wide range of applications such as urban planning，traffic navigation and virtual reality. The extraction of the geometric features(facades，windows and doors profiles) is a key step in the three-dimensional reconstruction of a building facade. In this paper，the planar structure of the building facade is first detected from point cloud data based on region growing and principal component analysis (PCA).Then，the boundaries of geometric features are defined by TIN contour extraction combined with a convex hull algorithm. Experiments show that the proposed method can effectively extract geometric features of the building facade from variable density point cloud data.