An automated classification system for urban areas matching the 'city country fingers' pattern: the cases of Kamakura (Japan) and Acireale (Italy) cities. Issue 1 (23rd September 2021)

• Record Type: Journal Article
• Authors: Spina, Roberto; Tramontana, Emiliano
• Is Part Of: Journal of urban ecology. Volume 7:Issue 1(2021)
• Publication Date: 2021-09-23
• Subjects: city fingers -- country fingers -- Alexander pattern; Urban ecology (Biology) -- Periodicals; 577.5605
• DOI: 10.1093/jue/juab023
• Languages: English
• ISSNs: 2058-5543

Abstract

Abstract: The uncontrolled expansion of urban areas is one of the main factors that reduce the liveability of cities. In recent years, to contrast urban sprawl, several nations have promoted policies aimed at developing urban green spaces. The importance of green oases within cities had already been highlighted, in 1977, by the architect Christopher Alexander who had developed a series of patterns including 'City Country Fingers' claiming that city development should consider the prolongation of country land in to the urban area. In several cities, especially in Japan, it is possible to recognize the imprint of urban development based on country fingers. This term refers to extensive urban intersections of agricultural land or wooded hills which, from the peripheral areas, penetrate the city. Inside them, there are urban windows, called city fingers, whose development direction is opposite to those of the country fingers. To recognize and analyze, in an automated way, these particular structures, a Python-based application was created. Starting from the original high-resolution image of Google Earth, a complete analysis was performed, labeling and delimiting urban and vegetational areas and extrapolating the main geometric parameters of the country and city fingers. The finalization of the results obtained was carried out through a classification model whose criteria were based on Alexander's pattern. Thanks to this classification scheme, the distinction between Active Green Areas (country fingers) and Passive Green Areas (gardens and public parks) have been revealed for the analyzed cities. The tests performed showed almost ideal conditions for the city of Kamakura and a limited match for the urban area of Acireale. The proposed method is suitable for fields of application that require a qualitative and quantitative determination of the vegetation cover present within the city, an essential condition for correct territorial planning.