A method for anomaly detection in hyperspectral images, using deep convolutional autoencoders

Menetelmä poikkeavuuksien havaitsemiseen hyperspektrikuvista käyttäen syviä konvolutiivisia autoenkoodereita. Poikkeavuuksien havaitseminen kuvista, erityisesti hyperspektraalisista kuvista, on hankalaa. Kun ongelmaan yhdistetään ennalta tuntematon data ja poikkeavuudet, muodostuu ongelma vielä...

Full description

Bibliographic Details
Main Author: Penttilä, Jeremias
Other Authors: Informaatioteknologian tiedekunta, Faculty of Information Technology, Informaatioteknologia, University of Jyväskylä, Jyväskylän yliopisto
Format: Master's thesis
Language:eng
Published: 2017
Subjects:
Online Access: https://jyx.jyu.fi/handle/123456789/55868
_version_ 1828193101207830528
author Penttilä, Jeremias
author2 Informaatioteknologian tiedekunta Faculty of Information Technology Informaatioteknologia University of Jyväskylä Jyväskylän yliopisto
author_facet Penttilä, Jeremias Informaatioteknologian tiedekunta Faculty of Information Technology Informaatioteknologia University of Jyväskylä Jyväskylän yliopisto Penttilä, Jeremias Informaatioteknologian tiedekunta Faculty of Information Technology Informaatioteknologia University of Jyväskylä Jyväskylän yliopisto
author_sort Penttilä, Jeremias
datasource_str_mv jyx
description Menetelmä poikkeavuuksien havaitsemiseen hyperspektrikuvista käyttäen syviä konvolutiivisia autoenkoodereita. Poikkeavuuksien havaitseminen kuvista, erityisesti hyperspektraalisista kuvista, on hankalaa. Kun ongelmaan yhdistetään ennalta tuntematon data ja poikkeavuudet, muodostuu ongelma vielä laajemmaksi. Spektraalisten poikkeavuuksien havaitsemiseen on kehitetty useita eri menetelmiä, mutta spatiaalisten poikkeavuuksien havaitseminen on huomattavasti hankalempaa. Tässä työssä esitellään uudenkaltainen menetelmä sekä spatiaalisten että spektraalisten poikkeavuuksien samanaikaiseen havaitsemiseen. Menetelmä on suunniteltu erityisesti spektraaliselle datalle, mutta soveltuu myös perinteisille kuville. Menetelmässä kolmiulotteisilla konvolutionaalisilla autoenkoodereilla löydetään koulutus-datassa esiintyviä normaaleja piirteitä. Tätä verkkoa käyttämällä voidaan testidata projisoida piirre-avaruuteen. Tästä projisoidusta datasta voidaan etsiä poikkeavuuksia käyttäen perinteisiä algoritmeja. Työssä esitetään kahdet erilliset tulokset. Ensimmäisissä on esitetty menetelmän toimivuus todellisuutta vastaavassa tilanteessa, jossa tietoa poikkeavuuksista ei ole etukäteen. Näiden tulosten lisäksi toinen ajo datalla, johon on lisätty synteettisiä tunnettuja poikkeavuuksia suoritetaan. Tämän toisen ajon tulokset voidaan validoida, koska anomaliat ovat nyt tunnettuja. Detecting anomalies from any image data, especially hyperspectral ones, is not a trivial task. When combined with the lack of apriori labels or detection targets, it grows even more complex. Detecting spectral anomalies can be done with numerous methods, but the detection of spatial ones is vastly more complicated affair. In this thesis a new way to detect both spatial and spectral anomalies at the same time is proposed. The method has been designed with hyperspectral data in mind, but should work for conventional images also. This is achieved works by using 3-d convolutional autoencoders to learn commonly occurring features both spatial and spectral, across the the test data. By running the test data through this network, the data is transformed to a feature-space. In this space, the images can be analyzed for the presence of anomalies by the means of standard anomaly detection algorithms. A simple real-world use case with unmodified images is presented. Second run for validation purposes is done with data containing synthetic anomalies.
first_indexed 2024-09-11T08:50:41Z
format Pro gradu
free_online_boolean 1
fullrecord [{"key": "dc.contributor.advisor", "value": "P\u00f6l\u00f6nen, Ilkka", "language": "", "element": "contributor", "qualifier": "advisor", "schema": "dc"}, {"key": "dc.contributor.advisor", "value": "H\u00e4m\u00e4l\u00e4inen, Timo", "language": "", "element": "contributor", "qualifier": "advisor", "schema": "dc"}, {"key": "dc.contributor.author", "value": "Penttil\u00e4, Jeremias", "language": null, "element": "contributor", "qualifier": "author", "schema": "dc"}, {"key": "dc.date.accessioned", "value": "2017-11-14T11:09:40Z", "language": "", "element": "date", "qualifier": "accessioned", "schema": "dc"}, {"key": "dc.date.available", "value": "2017-11-14T11:09:40Z", "language": "", "element": "date", "qualifier": "available", "schema": "dc"}, {"key": "dc.date.issued", "value": "2017", "language": null, "element": "date", "qualifier": "issued", "schema": "dc"}, {"key": "dc.identifier.other", "value": "oai:jykdok.linneanet.fi:1738562", "language": null, "element": "identifier", "qualifier": "other", "schema": "dc"}, {"key": "dc.identifier.uri", "value": "https://jyx.jyu.fi/handle/123456789/55868", "language": "", "element": "identifier", "qualifier": "uri", "schema": "dc"}, {"key": "dc.description.abstract", "value": "Menetelm\u00e4 poikkeavuuksien havaitsemiseen hyperspektrikuvista k\u00e4ytt\u00e4en syvi\u00e4\r\nkonvolutiivisia autoenkoodereita. \r\n\r\nPoikkeavuuksien havaitseminen kuvista, erityisesti hyperspektraalisista kuvista, on hankalaa. Kun ongelmaan yhdistet\u00e4\u00e4n ennalta tuntematon data ja poikkeavuudet, muodostuu ongelma viel\u00e4 laajemmaksi. Spektraalisten poikkeavuuksien havaitsemiseen on kehitetty useita eri menetelmi\u00e4, mutta spatiaalisten poikkeavuuksien havaitseminen on huomattavasti hankalempaa. T\u00e4ss\u00e4 ty\u00f6ss\u00e4 esitell\u00e4\u00e4n uudenkaltainen menetelm\u00e4 sek\u00e4 spatiaalisten ett\u00e4 spektraalisten poikkeavuuksien samanaikaiseen havaitsemiseen. Menetelm\u00e4 on suunniteltu erityisesti spektraaliselle datalle, mutta soveltuu my\u00f6s perinteisille kuville. Menetelm\u00e4ss\u00e4 kolmiulotteisilla konvolutionaalisilla autoenkoodereilla l\u00f6ydet\u00e4\u00e4n koulutus-datassa esiintyvi\u00e4 normaaleja piirteit\u00e4. T\u00e4t\u00e4 verkkoa k\u00e4ytt\u00e4m\u00e4ll\u00e4 voidaan testidata projisoida piirre-avaruuteen. T\u00e4st\u00e4 projisoidusta datasta voidaan etsi\u00e4 poikkeavuuksia k\u00e4ytt\u00e4en perinteisi\u00e4 algoritmeja. Ty\u00f6ss\u00e4 esitet\u00e4\u00e4n kahdet erilliset tulokset. Ensimm\u00e4isiss\u00e4 on esitetty menetelm\u00e4n toimivuus todellisuutta vastaavassa tilanteessa, jossa tietoa poikkeavuuksista ei ole etuk\u00e4teen. N\u00e4iden tulosten lis\u00e4ksi toinen ajo datalla, johon on lis\u00e4tty synteettisi\u00e4 tunnettuja poikkeavuuksia suoritetaan. T\u00e4m\u00e4n toisen ajon tulokset voidaan validoida, koska anomaliat ovat nyt tunnettuja.", "language": "fi", "element": "description", "qualifier": "abstract", "schema": "dc"}, {"key": "dc.description.abstract", "value": "Detecting anomalies from any image data, especially hyperspectral ones, is not a trivial task. When combined with the lack of apriori labels or detection targets, it grows even more complex. Detecting spectral anomalies can be done with numerous methods, but the detection of spatial ones is vastly more complicated affair. In this thesis a new way to detect both spatial and spectral anomalies at the same time is proposed. The method has been designed with hyperspectral data in mind, but should work for conventional images also. This is achieved works by using 3-d convolutional autoencoders to learn commonly occurring features both spatial and spectral, across the the test data. By running the test data through this network, the data is transformed to a feature-space. In this space, the images can be analyzed for the presence of anomalies by the means of standard anomaly detection algorithms. A simple real-world use case with unmodified images is presented. Second run for validation purposes is done with data containing synthetic anomalies.", "language": "en", "element": "description", "qualifier": "abstract", "schema": "dc"}, {"key": "dc.description.provenance", "value": "Submitted using Plone Publishing form by Jeremias Penttil\u00e4 (jefepent) on 2017-11-14 11:09:38.386140. Form: Pro gradu -lomake (https://kirjasto.jyu.fi/julkaisut/julkaisulomakkeet/pro-gradu-lomake). JyX data: [jyx_publishing-allowed (fi) =True]", "language": "en", "element": "description", "qualifier": "provenance", "schema": "dc"}, {"key": "dc.description.provenance", "value": "Submitted by jyx lomake-julkaisija (jyx-julkaisija.group@korppi.jyu.fi) on 2017-11-14T11:09:40Z\r\nNo. of bitstreams: 2\r\nURN:NBN:fi:jyu-201711144248.pdf: 9839130 bytes, checksum: 382c5fe91057460eed564c42908f9a11 (MD5)\r\nlicense.html: 4860 bytes, checksum: 6f72df65257a9cbaf939c0ba87a4373d (MD5)", "language": "en", "element": "description", "qualifier": "provenance", "schema": "dc"}, {"key": "dc.description.provenance", "value": "Made available in DSpace on 2017-11-14T11:09:40Z (GMT). No. of bitstreams: 2\r\nURN:NBN:fi:jyu-201711144248.pdf: 9839130 bytes, checksum: 382c5fe91057460eed564c42908f9a11 (MD5)\r\nlicense.html: 4860 bytes, checksum: 6f72df65257a9cbaf939c0ba87a4373d (MD5)\r\n Previous issue date: 2017", "language": "en", "element": "description", "qualifier": "provenance", "schema": "dc"}, {"key": "dc.format.extent", "value": "1 verkkoaineisto (75 sivua)", "language": null, "element": "format", "qualifier": "extent", "schema": "dc"}, {"key": "dc.format.mimetype", "value": "application/pdf", "language": null, "element": "format", "qualifier": "mimetype", "schema": "dc"}, {"key": "dc.language.iso", "value": "eng", "language": null, "element": "language", "qualifier": "iso", "schema": "dc"}, {"key": "dc.rights", "value": "In Copyright", "language": "en", "element": "rights", "qualifier": null, "schema": "dc"}, {"key": "dc.subject.other", "value": "hyperspektrikuvat", "language": null, "element": "subject", "qualifier": "other", "schema": "dc"}, {"key": "dc.subject.other", "value": "konvoluutio", "language": null, "element": "subject", "qualifier": "other", "schema": "dc"}, {"key": "dc.subject.other", "value": "autoenkooderit", "language": null, "element": "subject", "qualifier": "other", "schema": "dc"}, {"key": "dc.subject.other", "value": "machine learning", "language": null, "element": "subject", "qualifier": "other", "schema": "dc"}, {"key": "dc.subject.other", "value": "anomaly detection", "language": null, "element": "subject", "qualifier": "other", "schema": "dc"}, {"key": "dc.subject.other", "value": "hyperspectral images", "language": null, "element": "subject", "qualifier": "other", "schema": "dc"}, {"key": "dc.subject.other", "value": "hdbscan", "language": null, "element": "subject", "qualifier": "other", "schema": "dc"}, {"key": "dc.subject.other", "value": "convolutional neural network", "language": null, "element": "subject", "qualifier": "other", "schema": "dc"}, {"key": "dc.subject.other", "value": "autoencoder", "language": null, "element": "subject", "qualifier": "other", "schema": "dc"}, {"key": "dc.subject.other", "value": "convolutional autoencoder", "language": null, "element": "subject", "qualifier": "other", "schema": "dc"}, {"key": "dc.subject.other", "value": "CAE", "language": null, "element": "subject", "qualifier": "other", "schema": "dc"}, {"key": "dc.subject.other", "value": "SCAE", "language": null, "element": "subject", "qualifier": "other", "schema": "dc"}, {"key": "dc.subject.other", "value": "deep learning", "language": null, "element": "subject", "qualifier": "other", "schema": "dc"}, {"key": "dc.subject.other", "value": "autoenkooderi", "language": null, "element": "subject", "qualifier": "other", "schema": "dc"}, {"key": "dc.title", "value": "A method for anomaly detection in hyperspectral images, using deep convolutional autoencoders", "language": null, "element": "title", "qualifier": null, "schema": "dc"}, {"key": "dc.type", "value": "master thesis", "language": null, "element": "type", "qualifier": null, "schema": "dc"}, {"key": "dc.identifier.urn", "value": "URN:NBN:fi:jyu-201711144248", "language": null, "element": "identifier", "qualifier": "urn", "schema": "dc"}, {"key": "dc.type.ontasot", "value": "Pro gradu -tutkielma", "language": "fi", "element": "type", "qualifier": "ontasot", "schema": "dc"}, {"key": "dc.type.ontasot", "value": "Master\u2019s thesis", "language": "en", "element": "type", "qualifier": "ontasot", "schema": "dc"}, {"key": "dc.contributor.faculty", "value": "Informaatioteknologian tiedekunta", "language": "fi", "element": "contributor", "qualifier": "faculty", "schema": "dc"}, {"key": "dc.contributor.faculty", "value": "Faculty of Information Technology", "language": "en", "element": "contributor", "qualifier": "faculty", "schema": "dc"}, {"key": "dc.contributor.department", "value": "Informaatioteknologia", "language": "fi", "element": "contributor", "qualifier": "department", "schema": "dc"}, {"key": "dc.contributor.organization", "value": "University of Jyv\u00e4skyl\u00e4", "language": "en", "element": "contributor", "qualifier": "organization", "schema": "dc"}, {"key": "dc.contributor.organization", "value": "Jyv\u00e4skyl\u00e4n yliopisto", "language": "fi", "element": "contributor", "qualifier": "organization", "schema": "dc"}, {"key": "dc.subject.discipline", "value": "Tietotekniikka", "language": "fi", "element": "subject", "qualifier": "discipline", "schema": "dc"}, {"key": "dc.subject.discipline", "value": "Mathematical Information Technology", "language": "en", "element": "subject", "qualifier": "discipline", "schema": "dc"}, {"key": "dc.date.updated", "value": "2017-11-14T11:09:40Z", "language": "", "element": "date", "qualifier": "updated", "schema": "dc"}, {"key": "yvv.contractresearch.funding", "value": "0", "language": "", "element": "contractresearch", "qualifier": "funding", "schema": "yvv"}, {"key": "dc.type.coar", "value": "http://purl.org/coar/resource_type/c_bdcc", "language": null, "element": "type", "qualifier": "coar", "schema": "dc"}, {"key": "dc.rights.accesslevel", "value": "openAccess", "language": "fi", "element": "rights", "qualifier": "accesslevel", "schema": "dc"}, {"key": "dc.type.publication", "value": "masterThesis", "language": null, "element": "type", "qualifier": "publication", "schema": "dc"}, {"key": "dc.subject.oppiainekoodi", "value": "602", "language": null, "element": "subject", "qualifier": "oppiainekoodi", "schema": "dc"}, {"key": "dc.subject.yso", "value": "\u00e4lytekniikka", "language": null, "element": "subject", "qualifier": "yso", "schema": "dc"}, {"key": "dc.subject.yso", "value": "poikkeavuus", "language": null, "element": "subject", "qualifier": "yso", "schema": "dc"}, {"key": "dc.subject.yso", "value": "havaitseminen", "language": null, "element": "subject", "qualifier": "yso", "schema": "dc"}, {"key": "dc.subject.yso", "value": "neuroverkot", "language": null, "element": "subject", "qualifier": "yso", "schema": "dc"}, {"key": "dc.subject.yso", "value": "koneoppiminen", "language": null, "element": "subject", "qualifier": "yso", "schema": "dc"}, {"key": "dc.format.content", "value": "fulltext", "language": null, "element": "format", "qualifier": "content", "schema": "dc"}, {"key": "dc.rights.url", "value": "https://rightsstatements.org/page/InC/1.0/", "language": null, "element": "rights", "qualifier": "url", "schema": "dc"}, {"key": "dc.type.okm", "value": "G2", "language": null, "element": "type", "qualifier": "okm", "schema": "dc"}]
id jyx.123456789_55868
language eng
last_indexed 2025-03-31T20:03:10Z
main_date 2017-01-01T00:00:00Z
main_date_str 2017
online_boolean 1
online_urls_str_mv {"url":"https:\/\/jyx.jyu.fi\/bitstreams\/e5d3b79e-7a28-4e8a-889d-3893fb426cbe\/download","text":"URN:NBN:fi:jyu-201711144248.pdf","source":"jyx","mediaType":"application\/pdf"}
publishDate 2017
record_format qdc
source_str_mv jyx
spellingShingle Penttilä, Jeremias A method for anomaly detection in hyperspectral images, using deep convolutional autoencoders hyperspektrikuvat konvoluutio autoenkooderit machine learning anomaly detection hyperspectral images hdbscan convolutional neural network autoencoder convolutional autoencoder CAE SCAE deep learning autoenkooderi Tietotekniikka Mathematical Information Technology 602 älytekniikka poikkeavuus havaitseminen neuroverkot koneoppiminen
title A method for anomaly detection in hyperspectral images, using deep convolutional autoencoders
title_full A method for anomaly detection in hyperspectral images, using deep convolutional autoencoders
title_fullStr A method for anomaly detection in hyperspectral images, using deep convolutional autoencoders A method for anomaly detection in hyperspectral images, using deep convolutional autoencoders
title_full_unstemmed A method for anomaly detection in hyperspectral images, using deep convolutional autoencoders A method for anomaly detection in hyperspectral images, using deep convolutional autoencoders
title_short A method for anomaly detection in hyperspectral images, using deep convolutional autoencoders
title_sort method for anomaly detection in hyperspectral images using deep convolutional autoencoders
title_txtP A method for anomaly detection in hyperspectral images, using deep convolutional autoencoders
topic hyperspektrikuvat konvoluutio autoenkooderit machine learning anomaly detection hyperspectral images hdbscan convolutional neural network autoencoder convolutional autoencoder CAE SCAE deep learning autoenkooderi Tietotekniikka Mathematical Information Technology 602 älytekniikka poikkeavuus havaitseminen neuroverkot koneoppiminen
topic_facet 602 CAE Mathematical Information Technology SCAE Tietotekniikka anomaly detection autoencoder autoenkooderi autoenkooderit convolutional autoencoder convolutional neural network deep learning havaitseminen hdbscan hyperspectral images hyperspektrikuvat koneoppiminen konvoluutio machine learning neuroverkot poikkeavuus älytekniikka
url https://jyx.jyu.fi/handle/123456789/55868 http://www.urn.fi/URN:NBN:fi:jyu-201711144248
work_keys_str_mv AT penttiläjeremias amethodforanomalydetectioninhyperspectralimagesusingdeepconvolutionalautoencode AT penttiläjeremias methodforanomalydetectioninhyperspectralimagesusingdeepconvolutionalautoencoder