AN ORIGINAL ALGORITHM FOR BIM GENERATION FROM INDOOR SURVEY POINT CLOUDS
2019 ◽
Vol XLII-2/W13
◽
pp. 769-776
◽
Keyword(s):
3D Shape
◽
<p><strong>Abstract.</strong> Nowadays, it is essential to find new strategies, able to perform the first step of the scan-to-BIM process, by retrieving the geometrical information contained in point clouds that are so easily collected through laser scanners and range cameras. This paper presents a new algorithm for the automatic extraction of the layout and the height of a small indoor environment from its point cloud. In particular, the algorithm was tested on a point cloud of 600000 vertices, selected from the dataset of the ISPRS benchmark on indoor modelling. The preliminary results are encouraging: the 3D shape (layout and height) of the investigated room is effectively reconstructed.</p>