In this work, we present an innovative image recognition technique which is based on the exploitation of transit-data in images or simple photographs of sites of interest. Our objective is to automatically transform real-world images to graphs and, then, compute Minimum Spanning Trees (MST) in them.
We apply this framework and present an application which automatically computes efficient construction plans (for escalator or low-emission hot spots) for connecting all points of interest in cultural sites, i.e., archaeological sites, museums, galleries, etc, aiming to to facilitate global physical access to cultural heritage and artistic work and make it accessible to all groups of population.
