A Tool for Assessing Public Space Occupation as Applied to a Historical District: Alfama, Lisbon (Portugal)

Tourism is one of the sectors currently contributing the most to Portugal’s national economy. In 2017, the World Travel Awards elected Portugal as the World’s Leading Destination, and in 2019 Lisbon was awarded the Europe’s Leading City Tourist Board prize. According to the World Tourism Organization, more than 12 million tourists visited Portugal in 2018. Visitors’ dispersal has borne an effect on Lisbon’s historical neighbourhoods, but so far there is no way to measure how this affects community life, particularly regarding the use and occupation of the public space. This study aims to contribute to assessing public space occupation as based on image-processing techniques. For this purpose, a methodology has been developed for selecting the best location possible for installing cameras for the collection of images (based on GIS analysis and Space Syntax theory), and a software tool (TouriBeat Tool) to classify images according to public space occupation criteria. To carry out this research, the heart of the historical district, Alfama (Lisbon), was chosen as a case study, tourist and residential routes were identified in official guide maps, and person-to-person surveys were registered, analysed and mapped through GIS software. The listed routes were cross-checked with a syntactic analysis of the district’s urban network, enabling the identification of the most crowded and worrisome urban sockets as well as discussion of the best locations for measuring occupation rates. A strategic urban axis was selected for the installation of two cameras for image capturing (at 1 s acquisition rate). The images are automatically sent to a computer server for analysis by the software tool. A data set of the space percentage occupied by pedestrians is generated. The results delivered are visualised in a WEBSIG environment. The methodology proposed proved adequate to assess public space occupation: the use of low-cost equipment for data acquisition; fast and simple image processing techniques to classify images; and a user-friendly environment to display and analyse the results. Challenges arising from the presence of semi-permanent objects on the streets, as well as the solar angle conditions, have been addressed in the research.