Back to search

National Technical University of Athens (NTUA)

Χαρτογραφική γενίκευση φυσικών γραμμικών χωρικών οντοτήτων με δεσμεύσεις ευαναγνωσιμότητας

Abstract

dc:description

This thesis focuses on the research field of digital cartographic line generalization, by introducing a comprehensive method for generalizing natural linear entities. It addresses two key concepts of concern of the cartographic community in this field: 1. the generalization of linear entities 'guided' by the conceptual framework 'segmentation - characterization - generalization' and 2. the detection of critical points along cartographic lines.The thesis is developed at three levels: Firstly, a review of studies around this subject, which have been carried out, is presented. In particular, the concept of generalization of linear entities is introduced and current research trends in the field of cartographic generalization are listed and analyzed. The concept of critical points is introduced and their importance in the generalization process is highlighted. Then, there are presented the methods followed for automation in a digital environment, initially through simplification algorithms. The theoretical framework of operation of the most widespread of them is described and, based on the concept of critical points and the analysis of the relevant studies, an evaluation of them is been conducted. According to bibliography, the inadequacies of simplification algorithms lead researchers to look for new methods of generalizing cartographic lines. One of the innovative ideas being investigated is the identification of homogeneous lines or homogeneous segments along the lines, their classification according to qualitative and quantitative criteria, and their generalization via the relevant operators as appropriate.At the second level, two methodologies, that address the above defined questions are presented. Their structure and evaluation are developed in the main part of the dissertation. The first of these methodologies, aims to create a model of digital lines generalization, which is based on the conceptual framework 'segmentation-classification-generalization'. More specifically, a specific method has been designed, according to which each line is divided into homogeneous segments. The criteria for segmentation is the legibility of the lines, which is defined based on cartographic rules according to visual perception. Then, specific parameters of classification of the linear entity’s divided segments are introduced (the so called ε-segments) and four categories of segments (sections) are defined for classification, based on their shape and geometry. At the same time, quantitative metrics have been identified and calculated in order to define them. The generalisation of the line is completed by applying appropriate generalisation operators (merging, simplification, enlargement, etc.) to each part, depending on the section to which it belongs and its quantitative characteristics. For each segment generalization operators (enlargement, simplification, expansion, smoothing etc.) that are considered necessary for their generalization are proposed and their implementation algorithms are designed. The individual line processing operations are coded, interconnected and been integrated in an appropriate software environment so as to create a unified platform which will implement the techniques. The design of the method is carried out based on its implementation to lines representing physical entities, which, due to their complexity and randomness, are of particular research interest.The second method suggests a way of detecting the critical points of cartographic lines. The method is based on geometric principles and uses the length ratio index LR as a metric to estimate the slope change along a digital line. The LR index is applied, locally, to each point on the line and is structured in such a way that its value increases according to the change in slope of the line. The local examination is ensured by cutting the line through a circle of constant radius around each point of the line. The peaks of the line at which the index values show local maximum greater than a predetermined threshold are considered as critical points. In this document the structure of the LR indicator is analyzed and its operating parameters are defined. At the third level, a case study is carried out in order to test the effectiveness of the proposed methodologies. The critical points detection method is initially applied to four lines (three geomorphological and one geometrical), selected from two relevant studies, in order to be tested and evaluated from previously known critical points. The method is then piloted on the 1:50,000 scale coastline of the island of Peristera and the results are analysed. Finally, the LR index is used to compare the results of the implementation in five successive generalization scenarios of two line simplification algorithms ('pointremove' and 'bendsimplify' of ArcGIS software). The 'segmentation-classification-generalization' method is applied to the coastline of the island of Peristera at a scale of 1:50,000, for two generalisation scenarios with derived scales of 1:500,000 and 1:1,000,000. The objective is to evaluate both the individual processes and the final outcome resulting from the implementation of the technique to a random line of random shape and geometry and the scenarios of large scale change. The results are evaluated based on the predefined rules, but they are also compared with a) the results deriving from the implementation of two algorithms for line simplification ('pointremove' and 'bendsimplify') and b) the digitalization of analogue lines of GIS topographic maps at the respective scales.

Degree

thesis:*
Grantor dc:publisher
National Technical University of Athens (NTUA)
Year dc:date
2023

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Mitropoulos, Vasileios
  • Μητρόπουλος, Βασίλειος

Subjects

dc:subject × 16

Rights

Language dc:language
gre

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/53567

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Mitropoulos, Vasileios; Μητρόπουλος, Βασίλειος. Χαρτογραφική γενίκευση φυσικών γραμμικών χωρικών οντοτήτων με δεσμεύσεις ευαναγνωσιμότητας. National Technical University of Athens (NTUA), 2023. http://hdl.handle.net/10442/hedi/53567