Aristotle University Of Thessaloniki (AUTH)
Τεχνικές παραμορφώσιμων μοντέλων σε προβλήματα τεχνητής όρασης, επεξεργασίας εικόνας και βίντεο
Abstract
dc:descriptionThe main research of this PhD thesis are newly developed deformable techniques in computer vision problems, image and video processing. A new method for the automatic selection and tracking of characteristic feature points on 20 objects by using 3D deformable surfaces is presented. Additionally, a 1D and 2D discrete, non-separable signal transforms and their inverse formula are introduced. Moreover, a novel approach for estimating 3D head pose in single-new video sequences is proposed. Head pose is performed by using three appropriate trained radial basis function interpolation networks. Finally, a novel approach for image segmentation of color textured images, is presented 3D deformable surfaces, which combines image spatial information and image intensity, are exploited. The experimental results of all the aforementioned approaches show that they are useful techniques in computer vision and image and video processing.
Degree
thesis:*- Grantor dc:publisher
- Aristotle University Of Thessaloniki (AUTH)
- Year dc:date
- 2009
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Krinidis, Michail
- Κρηνίδης, Μιχαήλ
Subjects
dc:subject × 20- Παραμορφώσιμα μοντέλα επιφάνειας
- Παρακολούθηση χαρακτηριστικών σημείων
- Διακριτός μετασχηματισμός σήματος
- Συμπίεση εικόνας
- Προσανατολισμός ανθρώπινου προσώπου
- Δίκτυα παρεμβολής ακτινικών συναρτήσεων βάσης
- Κατάτμηση εικόνων
- Ψηφιακή επεξεργασία εικόνας
- Deformable surface models
- Feature tracking
- Discrete signal transformation
- Image compression
- Pose estimation
- Radial basis function interpolation
- Image segmentation
- Image processing
- Φυσικές Επιστήμες
- Επιστήμη Ηλεκτρονικών Υπολογιστών και Πληροφορική
- Natural Sciences
- Computer and Information Sciences
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/24400