University of Patras
Αναγνώριση αναδιπλώσεων με κάμερες βάθους και ρομποτικός χειρισμός υπερ-εύκαμπτων αντικειμένων
Abstract
dc:descriptionΤhe goal of this dissertation is the development of a complete solution for robotic unfolding, independent of the type, fabric, colors and shape of a garment as well as its decorative elements, such as pockets and collars. To achieve this, methods that incorporate human knowledge and perception are developed and combined with machine learning techniques. In addition, depth sensors are used, which provide more information about the garment’s configuration than a simple camera and are not affected by the colors of the garments or the shadows created by different lighting conditions.The robotic unfolding process followed in the proposed approach consists of three basic steps: 1) the garment is transformed from a random configuration, through robotic manipulations, to a half-folded state on a working table, 2) the configuration of the garment is recognized in order to ultimately extract its upper layer, 3) the garment is fully unfolded either by lifting it in the air from suitable points, or by rotating its upper layer around its folding axis.
Degree
thesis:*- Grantor dc:publisher
- University of Patras
- Year dc:date
- 2022
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Triantafyllou, Dimitra
- Τριανταφύλλου, Δήμητρα
Subjects
dc:subject × 14- Αναγνώριση αναδιπλώσεων
- Κάμερες βάθους
- Ρομποτικοί χειρισμοί
- Υπερ-εύκαμπτα αντικείμενα
- Fold detection
- Depth sensors
- Robotic manipulations
- Highly non-rigid objects
- Επιστήμη Ηλεκτρολόγου Μηχανικού, Ηλεκτρονικού Μηχανικού, Μηχανικού Η/Υ
- Επιστήμες Μηχανικού και Τεχνολογία
- Ρομποτική
- Electrical Engineering, Electronic Engineering, Information Engineering
- Engineering and Technology
- Robotics
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/51320