ACCESSIBLE PATH FINDING FOR HISTORIC URBAN ENVIRONMENTS: FEATURE EXTRACTION AND VECTORIZATION FROM POINT CLOUDS
Keywords: Semantic segmentation, Accessibility, Sidewalk inventory, Point cloud processing, Cultural heritage, Vectorization, QGIS, Network analysis
Abstract. Sidewalk inventory is a topic whose importance is increasing together with the widespread use of smart city management. In order to manage the city properly and to make informed decisions, it is necessary to know the real conditions of the city. Furthermore, when planning and calculating cultural routes within the city, these routes must take into account the specific needs of all users. Therefore, it is important to know the conditions of the city’s sidewalk network and also their physical and geometrical characteristics. Typically, sidewalk network are generated basing on existing cartographic data, and sidewalk attributes are gathered through crowdsourcing. In this paper, the sidewalk network of an historic city was produced starting from point cloud data. The point cloud was semantically segmented in ”roads” and ”sidewalks”, and then the cluster of points of sidewalks surfaces were used to compute sidewalk attributes and to generate a vector layer composed of nodes and edges. The vector layer was then used to compute accessible paths between Points of Interest, using QGIS. The tests made on a real case study, the historic city and UNESCO site of Sabbioneta (Italy), shows a vectorization accuracy of 98.7%. In future, the vector layers and the computed paths could be used to generate maps for city planners, and to develop web or mobile phones routing apps.