REMOTE SENSING AND URBAN METRICS: AN AUTOMATIC CLASSIFICATION OF SPATIAL CONFIGURATIONS TO SUPPORT URBAN POLICIES

The main aim of the research has been to create a classification system to evaluate the ecological characteristics of urban green spaces. From the current literature, we know that urban green allows us both to solve some problems such as air pollution, noise pollution and urban heat island, and to improve the aesthetic perception of urban quality. For this reason, in this paper the attention has been focused on the classification of urban green, and has been proposed an automatized methodology based on landscape metrics and remote sensing data, which have been distinguished homogeneous areas with the same structural and morphological characteristics of the vegetation. The innovative aspects of the method has been mainly three: the use of the coverage and height of the urban green from multispectral images and LiDAR data, the spatialization of this data on a central element of urban planning, namely city block, and, lastly, the use of a spatially bound geographic clustering procedure to identify both homogeneous and neighboring areas. The results have been shown the city of Livorno divided into different clusters, which have been highlighted the characteristics of the urban green. Each group has been characterized by city blocks that differ according to the shape and distribution of vegetation. The method permitted to differentiate the historical center from the suburbs and in according to this results we are able to determine guidelines useful for urban and landscape planning to regenerate and plan new urban spaces.