Master Thesis Template

Automatic detection and spatial quantification of trawl-marks in sidescan sonar images through image processing and analysis M A S T E R T H E S I S CHARIKLEIA GOURNIA April 2019 T h e s i s supervisor: Prof. Dr. George PAPATHEODOROU Environmental sciences School of natural sciences University of Patras Greece To George Papatheodorou who made this master’s study experience much more beautiful than my fantasy Supervisor professor: George Papatheodorou Examination Committee: Professor George Papatheodorou Professor Constantinos Koutsikopoulos Associate Professor Maria Geraga Επιβλέπων Καθηγητής: Γιώργος Παπαθεοδώρου Τριμελής Εξεταστική Επιτροπή: Καθηγητής Γιώργος Παπαθεοδώρου Καθηγητής Κωνσταντίνος Κουτσικόπουλος Αναπληρώτρια Καθηγήτρια Μαρία Γεραγά Abstract Bottom trawl footprints are a prominent environmental impact of deep-sea fishery that was revealed through the evolution of underwater remote sensing technologies. Image processing techniques have been widely applied in acoustic remote sensing, but accurate trawl-mark (TM) detection is underdeveloped. The paper presents a new algorithm for the automatic detection and spatial quantification of TMs that is implemented on sidescan sonar (SSS) images of a fishing ground from the Gulf of Patras in the Eastern Mediterranean Sea. This method inspects any structure of the local seafloor in an environmentally adaptive procedure, in order to overcome the predicament of analyzing noisy and complex SSS images of the seafloor. The initial preprocessing stage deals with radiometric inconsistencies. Then, multiplex filters in the spatial domain are performed with multiscale rotated Haar-like features through integral images that locate the TM-like forms and additionally discriminate the textural characteristics of the seafloor. The final TMs are selected according to their geometric and background environment features, and the algorithm successfully produces a set of trawling-ground quantification values that could be established as a baseline measure for the status assessment of a fishing ground. Keywords: Trawl marks; fishing grounds; automatic detection; side scan sonar; Haar- like features; morphological operations; seafloor characterization iii Περίληψη Η αλιεία με μηχανότρατες προκαλεί ένα πλήθος διαταράξεων στον πυθμένα της θάλασσας, εξέχουσες αυτών τα αλιευτικά ίχνη: γραμμικές ταπεινώσεις του πυθμένα που δημιουργούνται κατά τη σύρση των αλιευτικών εργαλείων. Η ανάπτυξη της τηλεπισκόπισης στο θαλάσσιο περιβάλλον φανέρωσε τους σχηματισμούς των αλιευτικών ιχνών των οποίων η ποσoτικοποίση αποτελεί βασικό μέτρο για την περιβαλλοντική παρακολούθηση των αλιευτικών πεδίων και την αξιολόγηση επιπτώσεων. Στόχος αυτής της εργασίας είναι η δημιουργία ενός αλγορίθμου ο οποίος επεξεργάζεται μεγάλο όγκο καταγραφών του ηχοβολιστικού συστήματος πλευρικής σάρωσης, εντοπίζει και ποσoτικοποιεί αυτόματα τα αλιευτικά ίχνη. Στην εργασία παρουσιάζεται επίσης η εφαρμογή του αλγορίθμου σε εικόνες που συλλέχθηκαν από ένα αλιευτικό πεδίο στον Κόλπο της Πάτρας. Οι εικόνες αρχίκα αναμορφώνονται με τεχνικές προεπεξεργασίας εικόνων, για να αναλυθούν κατά τον βέλτιστο τρόπο στο κύριο μέρος του αλγορίθμου. Στο κύριο μέρος της μεθόδου, μέσω χωρικών φίλτρων εικόνας βασισμένα σε χαρακτηριστικά τύπου Haar, εξετάζονται όλες οι δομές του πυθμένα σε μια διαδικασία που λαμβάνει υπόψη τους διαφορετικούς τύπους υφής του πυθμένα. Η πολυεπίπεδη ανάλυση χαρακτηριστικών στοχεύει στη διάκριση αλιευτικών ιχνών που παρουσιάζονται σε σύνθετα θαλάσσια περιβάλλοντα. Ανιχνευμένες δομές των οποίων τα φυσικά χαρακτηριστικά ταιριάζουν με την μορφή των αλιευτικών ιχνών και ανήκουν σε περιοχές με μορφολογία αντίστοιχη με αυτή που δημιουργούν τα αλιευτικά ίχνη καταμετρώνται και_ποσοτικοποιούνται. Λέξεις-κλειδιά: αλιευτικά ίχνη; αλιευτικά πεδία; αυτόματη αναγνώριση; ηχοβολιστής πλευρικής σάρωσης; χαρακτηριστικά τύπου Haar; μορφολογικές πράξεις iv Preamble Foreword This master thesis was conducted under the framework of the interdepartmental postgraduate course entitled "Environmental sciences" offered by the School of Natural Sciences of the University of Patras. The purpose of the thesis is to develop an image processing method that will automatically detect and quantify the footprints of trawl-fishing of large- scale areas that are mapped with sidescan sonar imaging system. The output data could be used as scientific database for environmental monitoring and management. Acknowledgments There aren't enough words to express how grateful I am that Elias Fakiris made me feel welcome in the laboratory of Marine Geology and Physical Oceanography of Department of Geology. He shared his knowledge with me and accompanied me till the end of my study, having common pursuit of approaching the logic. v Table of Contents 1 Introduction ................................................................................................................... 1 1.1 Monitoring of bottom-trawling impacts ......................................................... 1 1.2 Literature review .............................................................................................. 7 1.3 Objectives of this study ....................................................................................... 8 2 Study area and data pre-processing ............................................................................ 11 2.1 Study area ....................................................................................................... 11 2.2 Data acquisition and mapping ....................................................................... 13 2.3 Data interpretation ............................................................................................... 15 3 Methodology overview ............................................................................................... 16 3.1 Preprocessing techniques .................................................................................... 18 3.1.1 Intensity normalization .................................................................................. 18 3.1.2 Edge preserving smoothing ........................................................................... 20 3.2 Seafloor characterization and linear seabed feature detection ...................... 22 3.2.1 Multi-scale rotated Haar-like feature............................................................. 22 3.2.2 Seafloor characterization ............................................................................... 25 3.2.3 Linear seafloor features detection ................................................................. 27 3.2.4 Morphological operations .............................................................................. 29 3.3 TMs extraction and inclusion criteria ............................................................... 30 3.3.1 Geometrical criteria ....................................................................................... 30 3.3.2 Textural criteria ............................................................................................. 30 3.4 Trawling-grounds quantification ............................................................................ 31 4 Validation data and accuracy assessment ....................................................................... 32 5 Results .................................................................................................................................. 34 5.1 Trawl-marks extraction and quantification ............................................................. 34 5.2 Accuracy assessment .............................................................................................. 36 6 Discussion ............................................................................................................................. 40 7 Conclusion ........................................................................................................................... 45 8 References ........................................................................................................................... 48 Appendix................................................................................................................................. 63 Publication in journal and abstract for conference paper ................................................ 63 vi List of Figures Figure 1.1 Otter trawl nets and trawl doors ............................................................... 3 Figure 1.2 Satellite bottom trawling imagery ............................................................ 3 Figure 2.1 The study area ......................................................................................... 13 Figure 2.2 The SSS mosaic ...................................................................................... 14 Figure 3.1 Overall methodology workflow.............................................................. 18 Figure 3.2 Example image of image preprocessing filtering ................................... 21 Figure 3.3 Example image of image preprocessing filtering ................................... 22 Figure 3.4 Example image of image preprocessing filtering ................................... 22 Figure 3.5 The Haar-Like features ........................................................................... 24 Figure 3.6 Statistical measures over Haar-Like filter’s responses ........................... 25 Figure 3.7 Examples of Anisotropy and Complexity maps ..................................... 27 Figure 3.8 Example image of Trawl-marks enhancement filtering ........................

Load more