Back to search

National Technical University of Athens (NTUA)

Νέες προηγμένες μέθοδοι αναγνώρισης προτύπων με εφαρμογές στην επεξεργασία βιοσημάτων και την αυτόματη τρισδιάστατη ανακατασκευή θραυσμένων αντικειμένων

Abstract

dc:description

In the present thesis, new methodologies are presented in the fields of Pattern Analysis, Pattern Recognition and Curve and Surface Fitting. Each one of these methodologies supports the development of a corresponding information system for successfully tackling actual problems in the areas of Biomedicine and Archaeometry.Firstly, a new method is presented for processing bio-signals which belong to the class of Evoked Potentials. The study is based on the results of an experiment, which records the electroencephalic signals produced when a healthy subject is called upon to decide between a logical (Aristotelian) or paradoxical (according to Zeno) conclusion. Using advanced Pattern Recognition kai Signal Processing techniques, these bio-signals are classified in groups, according to the underlying brain functionality. Additionally, this thesis supports the assumption that logical (Aristotelian) deductions and Zeno’s paradoxes derive from different brain activities. The relative analysis strongly supports the argument that these activities may be based in different areas of the brain.Secondly, in the present work, new methods are presented for Pattern Recognition and Processing of 3-dimensional representations of fragments, aiming at developing an automated system for the determination of possible matches between fragments of any shape. Specifically, the application of the developed methodologies offered partial, virtual reconstruction of various fragmented archaeological findings of great historical significance. These findings include Mycenaean wall paintings belonging to the 13th century BC from the Tiryns excavation, as well as fragments of burial vessels of the same era, stored in the Prehistoric Collection of the National Archaeological Museum.Specifically, a first novel methodology has been developed for the reconstruction of fragmented objects, which have one side almost flat (as in the case of wall paintings), and a second one, for the reconstruction of fragmented objects of any geometrical form (as in the case of vessels). The introduced approach takes into consideration the 3-dimensional representations of the fragments, obtained via a 3-D scanner. Subsequently, the methodology uses a series of mathematically established criteria based on Calculus of Variations, which lead to the conclusion of whether two fragments match at a certain relative position. These criteria are either rejecting, in the sense that they conclude that two fragments do not match in a certain position, or sufficient, when they offer an irrevocable decision about matching.The application of these methods was accomplished through the development of corresponding original information systems, implementing the aforementioned criteria in a proper order. These systems offered islands of matching fragments belonging both to wall paintings and vessels. These islands of fragments have been spotted and virtually reconstructed for the first time.

Degree

thesis:*
Grantor dc:publisher
National Technical University of Athens (NTUA)
Year dc:date
2017

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Zannos, Solomon
  • Ζάννος, Σολομών

Subjects

dc:subject × 29

Rights

Language dc:language
gre

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/40410

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

Zannos, Solomon; Ζάννος, Σολομών. Νέες προηγμένες μέθοδοι αναγνώρισης προτύπων με εφαρμογές στην επεξεργασία βιοσημάτων και την αυτόματη τρισδιάστατη ανακατασκευή θραυσμένων αντικειμένων. National Technical University of Athens (NTUA), 2017. http://hdl.handle.net/10442/hedi/40410