A solution to the problem of the cartographic generalization of Italian geographical databases at large-medium scales: approach definition, process design and operators implementation

During many years in which the generalization of cartographic data has been studied many developments have been achieved. As some national mapping agencies in Europe and in the world are beginning to introduce automated processes in their production lines, the original dream of a completely automated system that could perform generalization is getting closer, even though it has not been reached yet. The aim of this dissertation is to investigate whether it is possible to design and implement a working generalization process for the Italian large-medium scale geographical databases. In this thesis we argue that the models, the approaches and the algorithms developed so far provide a robust and sound base to the problem of automated cartographic generalization, but that to build an effective generalization process it is necessary to deal with all the small details deriving from the actual implementation of the process on defined scales and data models of input and output. We propose/speculate that our goal can be reached by capitalizing the research results achieved so far and customizing the process on the data models and scales treated. This is the approach at the basis of this research work: the design of the cartographic generalization process and the algorithms implemented, either developed from scratch or deriving from previous works, have all been customized to solve a well defined problem: i.e. they expect input data that comply to a consistent data model and are tailored to obtain the results at defined scale and data model. This thesis explains how this approach has been brought into practice in the frame of the CARGEN project that aims at the development of a complete cartographic process to generalize the Italian medium scale geographical databases at 1:25000 and 1:50000 scale from the official Italian large scale geographical database at 1:5000 scale. This thesis will focus on the generalization to the 1:25000 scale, describing the approach that has been adopted, the overall process that has been designed and will provide details on the most important operators implemented for the generalization at such scale.

Questa tesi di dottorato sviluppa la problematica della generalizzazione cartografica applicata ai database geografici italiani alla alta e media scala. Il lavoro di ricerca si è sviluppato all'interno del progetto CARGEN, un progetto di ricerca tra l'Università di Padova e la Regione Veneto, con la collaborazione dell'IGMI per lo sviluppo di una procedura automatica di generalizzazione del database DB25 IGMI in scala 1:25000 a partire dal database regionale GeoDBR in scala 1:5000. Il lavoro di tesi affronta tutte le tematiche relative al processo di generalizzazione, partendo dalla generalizzazione del modello fino alla descrizione degli algoritmi di generalizzazione delle geometrie.