3D segmentation of unstructured point clouds for building modelling
P. Dorninger, and C. Nothegger. Photogrammetric Image Analysis (PIA07), 35, page 191--196. ISPRS, (2007)
Abstract
The determination of building models from unstructured three-dimensional point cloud data is often based on the piecewise intersection of planar faces. In general, the faces are determined automatically by a segmentation approach. To reduce the complexity of the problem and to increase the performance of the implementation, often a resampled (i.e. interpolated) grid representation is used instead of the original points. Such a data structure may be sufficient for low point densities, where steep surfaces (e.g. walls, steep roofs, etc.) are not well represented by the given data. However, in high resolution datasets with twenty or more points per square-meter acquired by airborne platforms, vertical faces become discernible making three-dimensional data processing adequate. In this article we present a three-dimensional point segmentation algorithm which is initialized by clustering in parameter space. To reduce the time complexity of this clustering, it is implemented sequentially resulting in a computation time which is dependent of the number of segments and almost independent of the number of points given. The method is tested against various datasets determined by image matching and laser scanning. The advantages of the three-dimensional approach against the restrictions introduced by 2.5D approaches are discussed.
%0 Conference Paper
%1 Dorninger2007
%A Dorninger, P.
%A Nothegger, C.
%B Photogrammetric Image Analysis (PIA07)
%D 2007
%E Stilla, Uwe
%E Mayer, Helmut
%E Rottensteiner, Franz
%E Heipke, Christian
%E Hinz, Stefan
%J International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences
%K Building Laserscanning segmentation 3DPointCloud Matching
%N 3/W49A
%P 191--196
%T 3D segmentation of unstructured point clouds for building modelling
%V 35
%X The determination of building models from unstructured three-dimensional point cloud data is often based on the piecewise intersection of planar faces. In general, the faces are determined automatically by a segmentation approach. To reduce the complexity of the problem and to increase the performance of the implementation, often a resampled (i.e. interpolated) grid representation is used instead of the original points. Such a data structure may be sufficient for low point densities, where steep surfaces (e.g. walls, steep roofs, etc.) are not well represented by the given data. However, in high resolution datasets with twenty or more points per square-meter acquired by airborne platforms, vertical faces become discernible making three-dimensional data processing adequate. In this article we present a three-dimensional point segmentation algorithm which is initialized by clustering in parameter space. To reduce the time complexity of this clustering, it is implemented sequentially resulting in a computation time which is dependent of the number of segments and almost independent of the number of points given. The method is tested against various datasets determined by image matching and laser scanning. The advantages of the three-dimensional approach against the restrictions introduced by 2.5D approaches are discussed.
@inproceedings{Dorninger2007,
abstract = {The determination of building models from unstructured three-dimensional point cloud data is often based on the piecewise intersection of planar faces. In general, the faces are determined automatically by a segmentation approach. To reduce the complexity of the problem and to increase the performance of the implementation, often a resampled (i.e. interpolated) grid representation is used instead of the original points. Such a data structure may be sufficient for low point densities, where steep surfaces (e.g. walls, steep roofs, etc.) are not well represented by the given data. However, in high resolution datasets with twenty or more points per square-meter acquired by airborne platforms, vertical faces become discernible making three-dimensional data processing adequate. In this article we present a three-dimensional point segmentation algorithm which is initialized by clustering in parameter space. To reduce the time complexity of this clustering, it is implemented sequentially resulting in a computation time which is dependent of the number of segments and almost independent of the number of points given. The method is tested against various datasets determined by image matching and laser scanning. The advantages of the three-dimensional approach against the restrictions introduced by 2.5D approaches are discussed.},
added-at = {2010-12-09T16:35:40.000+0100},
author = {Dorninger, P. and Nothegger, C.},
biburl = {https://www.bibsonomy.org/bibtex/2e116ef1594cb35f50a3530f19678de02/ipi_jn},
booktitle = {Photogrammetric Image Analysis (PIA07)},
editor = {Stilla, Uwe and Mayer, Helmut and Rottensteiner, Franz and Heipke, Christian and Hinz, Stefan},
file = {:\\\\james\\jukebox\\proceedings\\L_2007_PIA_MUNICH-PartA\\Papers\\PIA07_Dorninger_Nothegger.pdf:PDF},
interhash = {5187c11c03b618aa7ae08ce8e1a01ac8},
intrahash = {e116ef1594cb35f50a3530f19678de02},
journal = {International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences},
keywords = {Building Laserscanning segmentation 3DPointCloud Matching},
number = {3/W49A},
organization = {ISPRS},
pages = {191--196},
timestamp = {2010-12-09T16:35:40.000+0100},
title = {3D segmentation of unstructured point clouds for building modelling},
volume = 35,
year = 2007
}