Automatic Generation of Building Models in Dense Urban Areas Using Airborne Lidar and Aerial Photograph

Abstract: In this paper, an algorithm is proposed for automatically generating three-dimensional (3D) building models in dense urban areas. Automatic 3D building modeling in dense urban areas is challenging because, especially in Japan, houses that have slant roofs are located close to each other, and their heights are similar. For this case, difficulty in separating point clouds into individual buildings is an obstacle to modeling. To resolve this issue, the proposed algorithm uses the results of building segmentation from aerial photographs. Segmentation of buildings in urban areas, especially dense urban areas, by using remotely sensed images is also challenging because of the unclear boundaries between buildings and the shadows cast by neighboring buildings. The proposed algorithm successfully segments buildings from aerial photographs, including shadowed buildings in dense urban areas. The main factors in successful segmentation of shadowed roofs are (1) combination of different quantization results, (2) selection of buildings according to the rectangular index, and (3) edge completion by the inclusion of non-edge pixels that have a high probability of being edges. On the other hand, filtered airborne light detection and ranging (LiDAR) data are classified into small groups. By considering the segmented regions and the normals, models of actual building types—gable-roof, hip-roof, flat-roof and slant-roof buildings—are generated. To study the accuracy of the modeling, the proposed algorithm is applied to areas of Higashiyama ward, Kyoto, Japan. Owing to the information of building regions provided by segmentation, the modeling is successful even in dense urban areas. Therefore, the proposed algorithm is concluded to be effective in automatically generating building models in dense urban areas.