Outils pour la reconstruction automatique de bâtiments à partir d'imagerie aérienne

Christophe Vestri 1
1 ROBOTVIS - Computer Vision and Robotics
CRISAM - Inria Sophia Antipolis - Méditerranée
Abstract : Extracting building descriptions from an urban scene in 3D is an essential task for many industrial applications. Telecom Service providers require accurate Digital Elevation Models (DEM) for their planification tools which permit to deploy or to densify their cellular phone networks. With the objective of providing a product for telecom applications, this thesis focuses on building modeling. The main contributions of this thesis are the presentation, the study and the development of multiple tools for automatic building modeling. An original and important contribution is the automatic algorithm for reconstruction of the facades. It enables the reconstruction of facades without any a priori information about the orientation of the facades and the shape of the building. We propose to use the reconstructed points on the facades to correct the limits of the buildings in a DEM or to drive active contours. We also present an automatic modeling system of buildings based on the tools that we have proposed. This system of modeling is composed of the three following stages: (1) segmentation of the initial DEM in planar patches, (2) polygonal approximation of the contours of the planar patches for building an initial model and (3) orthogonalization of the building model. Finally, we propose to validate the results with telecom tools. Results seem satisfactory to deploy a mobile network in an rural or sub-urban area.
Document type :
Complete list of metadatas

Cited literature [90 references]  Display  Hide  Download

Contributor : Christophe Vestri <>
Submitted on : Thursday, December 11, 2008 - 2:49:00 PM
Last modification on : Saturday, January 27, 2018 - 1:30:45 AM
Long-term archiving on : Tuesday, June 8, 2010 - 4:26:35 PM


  • HAL Id : tel-00346358, version 1



Christophe Vestri. Outils pour la reconstruction automatique de bâtiments à partir d'imagerie aérienne. Autre. Université Nice Sophia Antipolis, 2000. Français. ⟨tel-00346358⟩



Record views


Files downloads