uberwachte-methoden-fur-die-spektrale-entmischung-mit-kunstlichen-neuronalen-netzen.pdf

In this work, artificial neural networks trained in a supervised manner for spectral unmixing are investigated. For this purpose, a suitable network architecture is determined first. After that, the focus lies on the generation of suitable training data. Model-based methods that generate training da...

Πλήρης περιγραφή

Λεπτομέρειες βιβλιογραφικής εγγραφής
Γλώσσα:German
Έκδοση: KIT Scientific Publishing 2023
Διαθέσιμο Online:https://doi.org/10.5445/KSP/1000159281
id oapen-20.500.12657-75890
record_format dspace
spelling oapen-20.500.12657-758902024-03-28T09:34:19Z Überwachte Methoden für die spektrale Entmischung mit künstlichen neuronalen Netzen Anastasiadis, Johannes data generation; data augmentation; supervised training; artificial neural network; hyperspectral image; Datenerzeugung; Datenaugmentierung; überwachtes Training; Hyperspektralbild; künstliche neuronale Netze In this work, artificial neural networks trained in a supervised manner for spectral unmixing are investigated. For this purpose, a suitable network architecture is determined first. After that, the focus lies on the generation of suitable training data. Model-based methods that generate training data from real pure spectra and data-based methods that augment existing training data are presented and evaluated. 2023-08-29T07:50:59Z 2023-08-29T07:50:59Z 2023 book https://library.oapen.org/handle/20.500.12657/75890 ger Forschungsberichte aus der Industriellen Informationstechnik application/pdf Attribution-ShareAlike 4.0 International uberwachte-methoden-fur-die-spektrale-entmischung-mit-kunstlichen-neuronalen-netzen.pdf https://doi.org/10.5445/KSP/1000159281 KIT Scientific Publishing 10.5445/KSP/1000159281 10.5445/KSP/1000159281 44e29711-8d53-496b-85cc-3d10c9469be9 29 198 open access
institution OAPEN
collection DSpace
language German
description In this work, artificial neural networks trained in a supervised manner for spectral unmixing are investigated. For this purpose, a suitable network architecture is determined first. After that, the focus lies on the generation of suitable training data. Model-based methods that generate training data from real pure spectra and data-based methods that augment existing training data are presented and evaluated.
title uberwachte-methoden-fur-die-spektrale-entmischung-mit-kunstlichen-neuronalen-netzen.pdf
spellingShingle uberwachte-methoden-fur-die-spektrale-entmischung-mit-kunstlichen-neuronalen-netzen.pdf
title_short uberwachte-methoden-fur-die-spektrale-entmischung-mit-kunstlichen-neuronalen-netzen.pdf
title_full uberwachte-methoden-fur-die-spektrale-entmischung-mit-kunstlichen-neuronalen-netzen.pdf
title_fullStr uberwachte-methoden-fur-die-spektrale-entmischung-mit-kunstlichen-neuronalen-netzen.pdf
title_full_unstemmed uberwachte-methoden-fur-die-spektrale-entmischung-mit-kunstlichen-neuronalen-netzen.pdf
title_sort uberwachte-methoden-fur-die-spektrale-entmischung-mit-kunstlichen-neuronalen-netzen.pdf
publisher KIT Scientific Publishing
publishDate 2023
url https://doi.org/10.5445/KSP/1000159281
_version_ 1799945243682930688