Back to search

Aristotle University Of Thessaloniki (AUTH)

Εξόρυξη πληροφορίας από 3d μοντέλα πόλης με τη χρήση γράφων για την υποστήριξη της διαδικασίας λήψης αποφάσεων στο χωρικό σχεδιασμό

Abstract

dc:description

The existence of a huge amount of information, which is available in the form of images and 3D models of the city, was the main motivational point for the preparation of this thesis dissertation. The main aim is to utilize this huge amount of information for the benefit of spatial planning. In recent decades, 3D city models seem to have been used primarily for visualization purposes. However, nowadays 3D city models have become very important tools that are used in a variety of applications in many different areas. The main objective of this research was to develop a methodology for the analysis of 3D city models and in particular the automatic clustering and classification of them, focusing on the successful identification of building structures belonging to a 3D city model. The detection of the building structures is carried out at the level of detail LOD1, which it focuses on building primitives identification. Hence, this thesis is aiming to develop a holistic open source tool that can support both the production of a 3D city model and its analysis. The tool which is implemented is pointed to be used and support the decision-making process in spatial planning. The proposed methodology (developed) for the automatic clustering and classification of the 3D city model is based on a data structure called graph along with the graph mining algorithms. The usage of the graph data structure for the buildings' identification of a 3D city model is the main novelty of this research. Most of the existing approaches are using supervised machine learning methods such as neural networks or deep learning techniques. This research proposes a non-supervised methodology by using the graph data structure. This approach has significant advantages such as the automation of building' identification process, the fast and efficient detection of the building structures, as shown by an extensive experimental process. Furthermore, the proposed methodology does not require the existence of a set of training data as in methodologies based on supervised learning methods. To sum up, the outputs of this thesis could be used for the benefit of spatial planning as an additional tool in the planning and decision-making process. The spatial planner engineer will be able to produce a 3D model of a city or a part of it that needs study. Then, by using the proposed methodology, the spatial engineer will be able to automatically detect the building structures that are located in the study area. After the successful identification of the building structures, the engineer will get as output many useful information like the number of building structures that are included in the 3D city model, buildings' heights as well as their position in the space. Thus, the proposed methodology and the tool will give to the spatial planner engineer a more realistic picture of the study environment and consequently helping him in the process of planning and decision making.

Degree

thesis:*
Grantor dc:publisher
Aristotle University Of Thessaloniki (AUTH)
Year dc:date
2021

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Valari, Eleni
  • Βαλάρη, Ελένη

Subjects

dc:subject × 16

Rights

Language dc:language
gre

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/49847

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

Valari, Eleni; Βαλάρη, Ελένη. Εξόρυξη πληροφορίας από 3d μοντέλα πόλης με τη χρήση γράφων για την υποστήριξη της διαδικασίας λήψης αποφάσεων στο χωρικό σχεδιασμό. Aristotle University Of Thessaloniki (AUTH), 2021. http://hdl.handle.net/10442/hedi/49847