Urban landscape assessment: a perceptual approach combining virtual reality and crowdsourced photo geodata

Abstract

The paper focuses on the evaluation of the visual quality of the cityscape of Livorno, in Italy. Using an algorithm based on Flickr’s Application Programming Interface, the coordinates of 7,453 shooting points of a shared photo were downloaded. A direct survey was conducted to assess the perception of urban landscape. The geographical spaces to be evaluated were measured using 78 spherical images played in virtual reality through a virtual reality headset.

The data were processed with the Principal Components Analysis. A Visual Quality Index (VQI) was developed to spatialise the results. Saliency maps were calculated to survey typical patterns in areas with high VQI. The results demonstrate how subjective visual perception helps identifying the criteria for the design of high-quality public spaces.