Klasifikacija hiperspektralnih snimaka metodom mapiranja spektralnog kuta i koeficijenta korelacije

master's thesis

• Title (english): Classification of Hyperspectral Imagery from the Air by the Spectral Angle Mapping and Correlation Coefficient Method
• Author: Filip Janeš
• Mentor: Jurica Ivošević (mentor)
• Committee member: Tomislav Radišić (predsjednik povjerenstva)
• Committee member: Doris Novak (član povjerenstva)
• Committee member: Tino Bucak (član povjerenstva)
• Granter: University of Zagreb; Faculty of Transport and Traffic Sciences; (Division of Aeronautics); (Department of Aeronautics); Zagreb
• Defense date and country: 2019-09-25, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Traffic and Transport Technology; Air Traffic

Abstract

Klasifikacija snimaka ima višestruku primjenu u području daljinskog istraživanja. Hiperspektralne snimke omogućuju najpreciznije spektralne karakteristike područja istraživanja. Izvršena je klasifikacija hiperspektralnih snimaka prirodnih minerala alunita, kaolinita i kalcedona u programu TNTmips. Definirane su spektralne vrijednosti pojedinih minerala te prikazana područja podudaranja uzoraka minerala sa okolnim područjima. Uporabom metoda mapiranja spektralnog kuta i koeficijenta korelacije pokazani su približno jednaki rezultati pokrivenosti minerala na područjima interesa. Preklapanjem rezultata pojedinih metoda, klasificirana područja su određena boljom preciznošću u usporedbi sa pojedinom metodom.

Abstract (english)

Image classification has multiple applications within the field of remote sensing. Hyperspectral images enable the most precise spectral characteristics inside the field of research. The hyperspectral image classification of natural minerals alunite, kaolinite and chalcedony was performed with the program TNTmips. Spectral values of individual minerals were defined and the fields of matching mineral samples with the adjecent areas were displayed. By using the spectral angle mapping and the correlation coefficient methods, the mineral coverage inside the areas of interest displayed similar results. By overlapping the results of each method, the classified fields are defined with better precision in comparison to each method individually.

• Keywords: klasifikacija snimke; hiperspektralna snimka; metoda mapiranja spektralnog kuta; metoda koeficijenta korelacije; spektralna vrijednost minerala
• Keywords (english): image classification; hyperspectral image; spectral angle mapping method; correlation coefficient method; mineral spectral value
• Language: croatian
• URN:NBN: urn:nbn:hr:119:653765
• Study programme: Title: Aeronautics; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka aeronautike (magistar/magistra inženjer/inženjerka aeronautike)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Open access
• Created on: 2019-11-15 12:34:18