Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XL-2/W3, 225-230, 2014
https://doi.org/10.5194/isprsarchives-XL-2-W3-225-2014
© Author(s) 2014. This work is distributed under
the Creative Commons Attribution 3.0 License.
 
22 Oct 2014
A DATA DRIVEN METHOD FOR BUILDING RECONSTRUCTION FROM LiDAR POINT CLOUDS
M. Sajadian and H. Arefi Dept. of Surveying and Geomatics Engineering, Faculty of Engineering, University of Tehran, Tehran, Iran
Keywords: 3D building model, Building extraction, Segmentation, Edge points detection, Line approximation Abstract. Airborne laser scanning, commonly referred to as LiDAR, is a superior technology for three-dimensional data acquisition from Earth's surface with high speed and density. Building reconstruction is one of the main applications of LiDAR system which is considered in this study. For a 3D reconstruction of the buildings, the buildings points should be first separated from the other points such as; ground and vegetation. In this paper, a multi-agent strategy has been proposed for simultaneous extraction and segmentation of buildings from LiDAR point clouds. Height values, number of returned pulse, length of triangles, direction of normal vectors, and area are five criteria which have been utilized in this step. Next, the building edge points are detected using a new method named "Grid Erosion". A RANSAC based technique has been employed for edge line extraction. Regularization constraints are performed to achieve the final lines. Finally, by modelling of the roofs and walls, 3D building model is reconstructed. The results indicate that the proposed method could successfully extract the building from LiDAR data and generate the building models automatically. A qualitative and quantitative assessment of the proposed method is then provided.
Conference paper (PDF, 2424 KB)


Citation: Sajadian, M. and Arefi, H.: A DATA DRIVEN METHOD FOR BUILDING RECONSTRUCTION FROM LiDAR POINT CLOUDS, Int. Arch. Photogramm. Remote Sens. Spatial Inf. Sci., XL-2/W3, 225-230, https://doi.org/10.5194/isprsarchives-XL-2-W3-225-2014, 2014.

BibTeX EndNote Reference Manager XML