Growing attention to sustainable mobility, the quality of public space, and the enhancement of historic heritage has made walkability a central criterion for assessing contemporary cities. The aim of this work is to propose an automated methodological procedure, oriented towards a value for people perspective, for the evaluation of urban walkability based on georeferenced street-level images. The methodology relies on the use of semantic segmentation techniques applied to open-source datasets (Google Street View and OpenStreetMap) and on the development of a synthetic index (Lvalue), capable of translating the visual perception of the urban environment into a quantitative value. The implementation of the procedure makes it possible to map the pedestrian quality of the street network, integrating infrastructural, environmental, and perceptual elements within a relational framework. The experimentation carried out in a densely built-up area of Campania produced high-resolution walkability maps, with evaluations referring to pedestrian accessibility thresholds (5, 10, and 15 minutes) to historic-cultural assets. The results show that the proposed approach allows for the identification of strengths and weaknesses of the pedestrian network at different spatial scales, highlighting local deficits, intermediate issues, and large-scale structural shortcomings. The procedure is configured as a decision-support tool for local administrations, capable of guiding targeted redevelopment interventions, enhancing cultural heritage, and monitoring the effects of urban policies. From this perspective, walkability understood as value for people emerges as a strategic indicator not only to promote more accessible and equitable cities but also to ensure lasting valorization of public space and historic heritage.
La crescente attenzione verso la mobilità sostenibile, la qualità dello spazio pubblico e la valorizzazione del patrimonio storico ha reso la walkability un criterio centrale per la valutazione delle città contemporanee. L’obiettivo del lavoro consiste nel proporre una procedura metodologica automatizzata, orientata al value for people, per la valutazione della camminabilità urbana a partire da immagini georeferenziate a livello stradale. La metodologia si fonda sull’utilizzo di tecniche di segmentazione semantica applicate a dataset open-source (Google Street View e OpenStreetMap) e sull’elaborazione di un indice sintetico (Lvalue), capace di tradurre la percezione visiva dell’ambiente urbano in un valore quantitativo. L’implementazione della procedura consente di mappare la qualità pedonale della rete viaria, integrando elementi infrastrutturali, ambientali e percettivi in una prospettiva relazionale. La sperimentazione condotta in un’area campana densamente urbanizzata ha prodotto mappe di walkability ad alta risoluzione, con valutazioni riferite a soglie temporali di accessibilità pedonale (5, 10 e 15 minuti) a beni storico-culturali. I risultati mostrano come l’approccio proposto permetta di individuare punti di forza e criticità della rete pedonale a diverse scale spaziali, evidenziando deficit locali, problematiche intermedie e carenze strutturali di ampio raggio. La procedura si configura come strumento di supporto decisionale per le amministrazioni locali, in grado di orientare interventi di riqualificazione mirata, valorizzare il patrimonio culturale e monitorare gli effetti delle politiche urbane. In tale prospettiva, la walkability intesa come value for people si afferma come indicatore strategico non solo per promuovere città più accessibili ed eque, ma anche per garantire una valorizzazione duratura dello spazio pubblico e del patrimonio storico.
Urban walkability assessment through shared images: a “value for people” approach to the public space
d'uva domenico;seccaroni marco
2026-01-01
Abstract
Growing attention to sustainable mobility, the quality of public space, and the enhancement of historic heritage has made walkability a central criterion for assessing contemporary cities. The aim of this work is to propose an automated methodological procedure, oriented towards a value for people perspective, for the evaluation of urban walkability based on georeferenced street-level images. The methodology relies on the use of semantic segmentation techniques applied to open-source datasets (Google Street View and OpenStreetMap) and on the development of a synthetic index (Lvalue), capable of translating the visual perception of the urban environment into a quantitative value. The implementation of the procedure makes it possible to map the pedestrian quality of the street network, integrating infrastructural, environmental, and perceptual elements within a relational framework. The experimentation carried out in a densely built-up area of Campania produced high-resolution walkability maps, with evaluations referring to pedestrian accessibility thresholds (5, 10, and 15 minutes) to historic-cultural assets. The results show that the proposed approach allows for the identification of strengths and weaknesses of the pedestrian network at different spatial scales, highlighting local deficits, intermediate issues, and large-scale structural shortcomings. The procedure is configured as a decision-support tool for local administrations, capable of guiding targeted redevelopment interventions, enhancing cultural heritage, and monitoring the effects of urban policies. From this perspective, walkability understood as value for people emerges as a strategic indicator not only to promote more accessible and equitable cities but also to ensure lasting valorization of public space and historic heritage.| File | Dimensione | Formato | |
|---|---|---|---|
|
siev.pdf
accesso aperto
:
Publisher’s version
Dimensione
2.35 MB
Formato
Adobe PDF
|
2.35 MB | Adobe PDF | Visualizza/Apri |
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.


