
Sidewalk Inventory in Sabbioneta
Managing the physical accessibility with point clouds and AI in an historic site
Scientific Director: Andrea Adami, Daniele Treccani
Collaborators: Lucía Díaz Vilariño
Partner: University of Vigo
Daniele Treccani, a PhD of Department ABC at Politecnico di Milano, and collaborator of He.Su.Tech. group, is studying in his research the management of physical accessibility in historical sites, in particular he is focusing on the AI methods for the automatic retrieval of useful attributes of sidewalks, basing on a point cloud. The title of his research is “Point Cloud Processing for the accessibility management in historic urban environments. The case of Unesco site of Sabbioneta.” and the main case study on which he is performing the study is Sabbioneta. The work was conducted together with Lucía Díaz Vilariño, a researcher from the University of Vigo. In his research, Daniele, developed, tested and exploited several methodologies that allows the automatic detection of sidewalks on a point cloud of an historic urban environment. The detected sidewalks are then characterized by geometric attributes such as the sidewalk width, its transverse and longitudinal slopes, the elevation respect the road and its paving material. The sidewalk network and the attributes are finally stored in a vector file that can be exploited by several finale users for many purposes. From this file several thematic maps were generated, to show the current values of sidewalks attributes. The generated shapefile could potentially be exploited form many purposes, from the management of accessibility and the plan of interventions, to the computation of routes within the city and the plan of visit paths for tourism purposes.

Attachments

Publications
- Sidewalk detection and pavement characterisation in historic urban environments from point clouds: preliminary results
- A data collection framework for managing accessibility and inclusion in urban heritage
- Accessible path finding for historic urban environments: feature extraction and vectorization from point clouds
- A deep learning approach for the recognition of urban ground pavements in historical sites
architectural preservation and valorization environment and landscape geomatics inclusion and accessibility