The International Association for Automation and Robotics in Construction
Publications / 2011 Proceedings of the 28th ISARC, Seoul, Korea
An Approach to Extracting Facade Features Using the Intensity Value of Terrestrial Laser Scanner with Mathematical Morphology
Nowadays, Terrestrial Laser Scanner (TLS) is a crucial device in order to acquire geographical data easily. In recent years, advances in accuracy and velocity have enhanced 3D models, but these stand-alone data do not provide the feature extraction. High density point clouds can be obtained with terrestrial laser scanner and by means of modelling and processing techniques, objects features can be constructed. Generally researchers use the geometric information for extracting object characteristics. This paper describes a methodology to use not only the geometrical information but also the intensity data provided during the scan process. The aim of this paper is to present a methodology to extract façades from laser scanner data. The proposed method uses mathematical morphology to classify, segment and extract the borders of principal features of the building Quiñones de León country mansion in the city of Vigo, Galicia, Spain. Furthermore, the final result shows huge potential of the intensity data solution to extract the main features of façades and together with the K-means algorithm for segmentation and mathematical morphology operators. It suggests keeping the effort going on combining them as successful tools employed for feature extraction of historical buildings.