OBTAINING 3D SEMANTIC OBJECTS FOR URBAN APPLICATIONS
The main objective of the SEM3D project was to extract objects from the urban fabric using artificial intelligence methods from the new airborne LIDAR data for the Walloon region and the corresponding aerial photos. The working area for the implementation of the POC (proof-of-concept) is the Outremeuse district. The results of the semantic segmentation were used to provide geographical information at a very high level of detail, including infrastructure, trees, transport networks and buildings. The semantic and geometric richness of this information should make it possible to improve simulation systems and tools, as well as the analysis of key urban issues.
