Results - Details

Search command : Author="Τζιρίτας"  And Author="Γεώργιος"

Current Record: 8 of 12

[Add to Basket]
Identifier	000382630
Title	Τεχνικές ανάλυσης εικόνων για την ανίχνευση δομικών στοιχείων σε προσόψεις κτιρίων
Alternative Title	Image analysis techniques for the detection of structural elements in building facades
Author	Ζωράκης, Νικόλαος
Thesis advisor	Τζιρίτας, Γεώργιος
Reviewer	Αργυρός, Αντώνιος Παναγιωτάκης, Κωνσταντίνος
Abstract	The goal of this work is the structure detection in building façade images. We developed techniques that start with a segmented facade image. The target is the classification of the individual regions from which the facade image is composed. Relevant descriptive geometrical features are extracted from the detected regions that are used for the determination of the structure that belong to. Based on the features that are extracted from a collection of building facade images of similar architectural design with supervised learning, the representative descriptive values of structural elements are specified, so as to be used to the final stage of the structure recognition. At the final phase, they are proposed and used recognition techniques of the sky, the wall and the windows. Special focus is given to the recognition of areas that correspond to windows, by developing methods that are based on the characteristics of these areas, the periodicity, the symmetry and the fundamental architectural principles of the building facades. At first, the recognition is done depending only to the characteristics with high degree of likelihood. Subsequently, the properties of the periodicity and the architectural structure of the facade are exploited. We present results from different image collections with various characteristics. The final recognition result is considered good, on the premise that the segmentation of building facade image is satisfactory.
Language	Greek
Issue date	2014-03-28
Collection	School/Department--School of Sciences and Engineering--Department of Computer Science--Post-graduate theses
	Type of Work--Post-graduate theses
Permanent Link	https://elocus.lib.uoc.gr//dlib/1/6/d/metadata-dlib-1392710385-213859-4135.tkl
Views	590

Digital Documents
	Download document View document Views : 19