Aristotle University Of Thessaloniki (AUTH)
Μοντέλα ανάλυσης τηλεπισκοπικών δεδομένων για τον εντοπισμό και την παρακολούθηση μεταβολών σε αστικές περιοχές
Abstract
dc:descriptionThe current remotely sensed data triggered the development of novel applications enabling the monitoring of urban areas. Nowadays digital image processing is focusing on the development of algorithms for the effective mapping of the dynamic urban environment. This tendency is reinforced by the research interest of the National Mapping and Cadastral Agencies in most European countries towards the implementation of strategies for the updating of 2D topographic databases. The proposed multitemporal image analysis combines the advantages of both pixel- and object-based image analysis. Change detection algorithms are applied on the simplified images for the detection of altered image pixels. Object-based image analysis is subsequently introduced for the efficient handling of changes by implementing a knowledge-based classification scheme. The developed algorithm is designed to provide a solution even when only optical satellite imagery is available. A second alternative change detection approach is established, diverging from the previous in a way that it exploits vector information of an existing geodatabase along with imagery data. A knowledge-base model is designed for the extraction of the objects of interest with the support of geodatabase information. The quantitative and qualitative evaluation of the proposed methodologies is performed with the help of ancillary ground truth data. The quantitative evaluation is accomplished by employing the standard performance evaluation measures of detection completeness, correctness and overall quality, which have been widely applied on building identification studies. The experimental results and the performed quantitative evaluation demonstrate the potential of the developed knowledge-based classification frameworks. The complexity of urban features dictates the combination of spectral, geometric and topological information under an efficient object-oriented analysis.
Degree
thesis:*- Grantor dc:publisher
- Aristotle University Of Thessaloniki (AUTH)
- Year dc:date
- 2011
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Δοξάνη, Γεωργία
- Doxani, Georgia
Subjects
dc:subject × 16- Τηλεπισκόπηση
- Διαχρονικές εικόνες
- Εντοπισμός μεταβολών
- Φίλτρα χώρου - κλίμακας
- Αντικειμενοστραφής ανάλυση εικόνας
- Ταξινόμηση βασισμένη στη γνώση
- Remote sensing
- Multitemporal images
- Change detection
- Scale - space filtering
- Object - based image analysis
- Knowledge - based classification
- Επιστήμες Μηχανικού και Τεχνολογία
- Επιστήμη Πολιτικού Μηχανικού
- Engineering and Technology
- Civil Engineering
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/26979