diplomsko delo
Marko Kužner (Author), Domen Mongus (Mentor), David Selčan (Co-mentor), Tomaž Rotovnik (Co-mentor)

Abstract

V tem diplomskem delu predstavljamo analizo metod za zaznavanje oljnih madežev na vodni površini s satelitom TRISAT. Primerjali smo metodo največjega verjetja in nevronsko mrežo. Algoritma smo učili in testirali nad dvema različnima bazama podatkov. Z rezultati smo pokazali, da je metoda največjega verjetja računsko in prostorsko bolj spremenljiva pri manjšem številu vhodnih podatkov, medtem ko se je nevronska mreža izkazala za natančnejšo. S primerjavo najboljših izbranih kanalov nad bazama podatkov smo pokazali, da so si izbrani kanali podobni. Rezultate tega diplomskega dela lahko uporabimo za izvedbo algoritma nad referenčnimi slikami satelita TRISAT.

Keywords

satelit TRISAT;metoda največjega verjetja;nevronske mreže;zaznavanje olja;kratkovalovni infrardeči spekter;diplomske naloge;

Data

Language: Slovenian
Year of publishing:
Typology: 2.11 - Undergraduate Thesis
Organization: UM FERI - Faculty of Electrical Engineering and Computer Science
Publisher: [M. Kužner]
UDC: 004.93:629.783(043.2)
COBISS: 41195267 Link will open in a new window
Views: 683
Downloads: 0
Average score: 0 (0 votes)
Metadata: JSON JSON-RDF JSON-LD TURTLE N-TRIPLES XML RDFA MICRODATA DC-XML DC-RDF RDF

Other data

Secondary language: English
Secondary title: Detection of oil spills in multispectral satellite images
Secondary abstract: This diploma thesis explored methods for the detection of oil spills in multispectral satellite images. We compared the maximum likelihood classification and neural networks. Algorithms were trained and tested over two different databases. Results showed that the maximum likelihood classification is computationally and space complexity more suitable for a smaller number of inputs, while the neural network proved to be more accurate on the other side. A comparison of the best selected channels over the databases showed that selected channels were similar. The results of this diploma thesis can be used for the implementation of an algorithm on the reference images of satellite TRISAT.
Secondary keywords: satellite TRISAT;maximum likelihood classification;neural networks;oil detection;short-wavelength infrared spectrum;
Type (COBISS): Bachelor thesis/paper
Embargo end date (OpenAIRE): 2023-08-31
Thesis comment: Univ. v Mariboru, Fak. za elektrotehniko, računalništvo in informatiko, Računalništvo in informacijske tehnologije
Pages: X, 30 f.
ID: 11995567