Building Feature Extraction from Airborne Lidar Data Based on Tensor Voting Algorithm

This study presents a novel approach based on the tensor voting framework for extracting building features from airborne lidar data. Geometric features of lidar points are represented by a tensor field in this paper. For the extraction of roof patches, a region-growing method with principal features is developed from the properties of eigenvalues and eigenvectors of the tensor field. Additionally, three new indicators for the strength of features are presented to reduce the effect of the number of points on feature identification, and a supervised method is proposed to determine the threshold of planar feature strength for the region-growing. The extraction of ridge and edge lines from the segmented roof patches is also discussed. Experiments based on airborne lidar data are described to demonstrate the effectiveness of the proposed method, with those the results compared with the PCA method.

[1]  Robert C. Bolles,et al.  Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography , 1981, CACM.

[2]  Robert De Wulf,et al.  Sub-pixel mapping with neural networks : real-world spatial configurations learned from artificial shapes , 2003 .

[3]  B. Gorte.  SEGMENTATION OF TIN-STRUCTURED SURFACE MODELS , 2002 .

[4]  George Vosselman,et al.  Two algorithms for extracting building models from raw laser altimetry data , 1999 .

[5]  Norbert Pfeifer,et al.  A Comprehensive Automated 3D Approach for Building Extraction, Reconstruction, and Regularization from Airborne Laser Scanning Point Clouds , 2008, Sensors.

[6]  Ruijin Ma.  DEM Generation and Building Detection from Lidar Data , 2005 .

[7]  Miao Wang,et al.  Automatic plane extraction from LIDAR data based on octree splitting and merging segmentation , 2005, Proceedings. 2005 IEEE International Geoscience and Remote Sensing Symposium, 2005. IGARSS '05..

[8]  E. Zelniker,et al.  Detection and vectorization of roads from lidar data , 2007 .

[9]  J. R. Jensen,et al.  Creation of digital terrain models using an adaptive lidar vegetation point removal process , 2002 .

[10]  J. Shan,et al.  Building boundary tracing and regularization from airborne lidar point clouds , 2007 .

[11]  Cheng Wang,et al.  Separation of Ground and Low Vegetation Signatures in LiDAR Measurements of Salt-Marsh Environments , 2009, IEEE Transactions on Geoscience and Remote Sensing.

[12]  Joonki Paik,et al.  Normal Vector Voting: Crease Detection and Curvature Estimation on Large, Noisy Meshes , 2002, Graph. Model..

[13]  E. Næsset,et al.  Single Tree Segmentation Using Airborne Laser Scanner Data in a Structurally Heterogeneous Spruce Forest , 2006 .

[14]  E. Kjems,et al.  Automatic 3D building reconstruction from airbornelaser scanning and cadastral data using hough transform , 2004 .

[15]  T. Soler,et al.  On covariances of eigenvalues and eigenvectors of second-rank symmetric tensors , 1991 .