1. Αποθετήριο Παραδοτέων Υποτρόφων Δράσεων ΕΣΠΑ Τριτοβάθμιας Εκπαίδευσης 2017-2023
2. Όλο το Αποθετήριο
3. Ερευνητικές ομάδες

Τύπος:	Αναρτημένη ανακοίνωση (poster)
Τίτλος:	Semi-automatic extraction of stream bed grain-size classes based on UAS derived data
Συγγραφέας:	[EL] Μαρκογιάννη, Βασιλική[EN] Markogianni, Vassiliki [EL] Παπαϊωάννου, Γεώργιος[EN] Papaioannou, Georgios [EL] Λουκάς, Αθανάσιος[EN] Loukas, Athanasios [EL] Δημητρίου, Ηλίας[EN] Dimitriou, Elias
Ημερομηνία:	16/12/2021
Περίληψη:	A significant and challenging issue in river geomorphological research including sediment transport, hydraulic resistance, and the prediction of flow velocity, is the quantification of the river bed grain sizes. Since field sampling methods provide only pointwise information, a more representative characterization of inhomogeneous bed composition is needed, acquired in a time efficient manner. This study presents a semi-automatic methodοlogy to gain areal information of surface grain size classes using airborne imagery and ground truth data, sampled from 11 grids (1x1 m2) equally distributed across the stream. High-resolution RGB orthomosaic images with resolutions of 1.34 cm/px were generated from RGB images acquired with an Unmanned Aerial Vehicle (UAV). A variety of pixel-based and object-based image analyses were examined to acquire the most accurate classification. The examined river bed is not the typical case (e.g. gravel dominated, rounded particles) and consists of a mixed gravel and cobble bed material with sharp-edges. Eventually, object-based image classification and particularly a grey-level co-occurrence matrix (GLCM) was used to examine several texture parameters. Local entropy values in combination with Maximum Likelihood Classifier (MLC; pixel-based unsupervised classification method) were highlighted as a satisfactory approach to determine all existing grain classes along the stream bed. The methodology is demonstrated at the lower part of Xerias stream reach (2.2 km), Volos, Greece. The qualitative capability of the developed hybrid method to classify and map the stream bed was evaluated in two ways. Initially, 240 random points were created and equally distributed among the three known grain classes (boulder, cobble, gravel) according to field measurements. This methodology indicated that an overall 65% correct classification was achieved. Then, considering the grain shape assessment of the field measurements, the typical sediment area shape formulas were used to estimate the area percentage of each grain class for each grid. Comparing the estimated area percentage based on the field measurements with the generated surface grain size classified map provided an average overall 52% correct classification.
Γλώσσα:	Αγγλικά
Τόπος δημοσίευσης:	New Orleans, U.S.A
Σελίδες:	1
Θεματική κατηγορία:	[EL] Άλλες φυσικές επιστήμες[EN] Other natural sciences
Κάτοχος πνευματικών δικαιωμάτων:	© 2021 The Author(s)
Ηλεκτρονική διεύθυνση του τεκμηρίου στον εκδότη:	https://agu.confex.com/agu/fm21/meetingapp.cgi/Paper/859147
Ηλεκτρονική διεύθυνση περιοδικού:	https://agu.confex.com/agu/fm21/meetingapp.cgi/Home/0
Όνομα εκδήλωσης:	AGU Fall Meeting 2021
Τοποθεσία εκδήλωσης:	New Orleans, LA and Online Everywhere
Ημ/νία έναρξης εκδήλωσης:	13/12/2021
Ημ/νία λήξης εκδήλωσης:	17/12/2021
Σημειώσεις:	“This research is co-financed by Greece and the European Union (European Social Fund- ESF) through the Operational Programme «Human Resources Development, Education and Lifelong Learning 2014-2020» in the context of the project “Remote sensing methodology for roughness estimation in ungauged streams and sensitivity analysis of floods using different hydraulic/hydrodynamic modeling approaches (1D,2D,1D/2D) (MIS 5048553).”
Εμφανίζεται στις συλλογές:	Ερευνητικές ομάδες