Aristotle University Of Thessaloniki (AUTH)
Ανάπτυξη αυτοματοποιημένων μεθόδων τηλεπισκόπησης για την εκτίμηση ξυλώδους όγκου δασικών οικοσυστημάτων
Abstract
dc:descriptionKnowledge of forest parameters is essential for monitoring and management of forest ecosystems. It is important at a regional scale (i.e. due to forest management plans) as well as at a global scale (i.e. due to climate change). The evolution in the sector of satellite imaging systems, commercial software processing and analysis of remote sensing images and Geographic Information Systems have provided new opportunities in research. In this context, the development of automated remote sensing methods was attempted for calculating parameters in order to estimate wood volume of forest stand trees. The proposed methodology was developed and evaluated in study areas considering the most important Greek tree species. Therefore, Mediterranean forest ecosystems, being the wider geographical area of this research, and the main Greek forest tree species were presented. In the up to now wood volume estimation approaches, field measurements have been extensively used. This creates the premise and the restriction of field data in order to estimate wood volume. As a result, e.g. wood volume cannot be assessed in inaccessible areas, because field data cannot be collected. The aforementioned facts gave the motivation for the development of automated methods in this research, which help in the improvement and the strengthening of others’ weak points. The results of the required bibliographical research were presented thoroughly. In the survey for the development of these methods, a theoretical investigation for the use of remote sensing stereo images was conducted. An evaluation of the Cartosat-1 stereoimage usability in forest tree height information extraction was carried out. Height information extraction is quite expensive while requires the processing of other data types in order to acquire the necessary accuracy and detail (active remote sensing systems), making the procedure more complicated. Therefore, particular emphasis was given to the assessment of wood volume using monoscopic satellite images. The development of the proposed methodology was performed using a Quickbird-2 image, with a spatial resolution of 0.6m in the Pentalofo forest area (Kozani). The evaluation of the results was based on accurate and detailed field data, recorded by the Tsotyli Forest Service. The tree species under study was black pine (Pinus nigra). In order to generally evaluate the proposed methodology, it was applied in other study areas, in different forests, and on different tree species. More specifically the tree species were calabrian pine (Pinus brutia) and oak (Quercus conferta), situated in the Taxiarchis and Palaiokastro forest regions. In this way, the transferability in terms of geographical area and tree species was examined. Moreover, during the evaluation, a different Quickbird-2 image was utilized. The results of the proposed methodology, developed in this dissertation, were very encouraging and show its potential.
Degree
thesis:*- Grantor dc:publisher
- Aristotle University Of Thessaloniki (AUTH)
- Year dc:date
- 2015
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Stournara, Panagiota
- Στουρνάρα, Παναγιώτα
Subjects
dc:subject × 18- Τηλεπισκόπηση
- Αντικειμενοστραφής ανάλυση εικόνας
- Δορυφορικές εικόνες
- Ξυλώδης όγκος
- Δασικά οικοσυστήματα
- Remote sensing
- GEOBIA (Geographic-Object-Based Image Analysis)
- Satellite imagery
- Wood volume
- Forest ecosystems
- Επιστήμες Μηχανικού και Τεχνολογία
- Άλλες Επιστήμες Μηχανικού και Τεχνολογίες
- Γεωπονικές Επιστήμες και Κτηνιατρική
- Γεωπονία, Δασολογία και Αλιεία
- Engineering and Technology
- Other Engineering and Technologies
- Agricultural and Veterinary Sciences
- Agriculture, Forestry and Fisheries
Rights
- Language dc:language
- gre
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/35401