Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • harvard-cite-them-right
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf
Objektdetektering med hjälp av maskininlärning och neurala nätverk
University of Borås, Faculty of Textiles, Engineering and Business.
University of Borås, Faculty of Textiles, Engineering and Business.
University of Borås, Faculty of Textiles, Engineering and Business.
2020 (Swedish)Independent thesis Basic level (university diploma), 180 HE creditsStudent thesisAlternative title
Object detection using machine learning and neural networks (English)
Abstract [sv]

I denna rapport presenteras utvärdering av prestanda av tre maskinlärningsalgoritmer för objektdetektering. Studiens syfte är att öka kunskap inom maskininlärning genom att skapa förståelse för hur objektdetektering fungerar. Metoden som valdes var att först få en uppfattning om ämnet genom att studera berörande litteratur och därefter utföra den praktiska jämförelsestudien.

Arbetet behandlar frågor och ger en uppfattning om hur objektdetektering fungerar och hur maskininlärning används. I resultatet visades Convolutional Neural Network vara överlägsen jämfört med Artificial Neural Network och Haar Cascade gällande objektdetektering.

Jämförelsen mellan Artificial Neural Network och Convolutional Neural Network med samma aktiveringsfunktion gav att Convolutional Neural Network hade bättre klassificerings noggrannhet. Aktiveringsfunktionen ReLU visades vara den bästa kombinationen i samband med Arificial Neural Network och Convolutional Neural Network.

Jämförelsestudien mellan algoritmerna visar att Convolutional Neural Network uppnår bäst resultat både när det gäller detektering av siffror och ansikten.

Abstract [en]

This report presents surveys of three different techniques for object detection. The purpose of the report is to increase knowledge in machine learning by creating an understanding of how object detection works. The chosen method is first to get an idea of the subject by studying the relevant literature and then perform the practical comparative studies.

The work addresses issues and gives an idea of how object detection works and how machine learning is used. In the result, Convolutional Neural Network was shown to be superior to Artificial Neural Network and Haar Cascade in object detection.

The comparison between the Artificial Neural Network and the Convolutional Neural Network with the same activation function gave the Convolutional Neural Network better classification accuracy.

The activation function ReLU was shown to be the best combination in conjunction with the Arificial Neural Network and the Convolutional Neural Network. Comparative studies between the algorithms show that the Convolutional Neural Network achieves the best results in both figure and face detection.

Place, publisher, year, edition, pages
2020.
Keywords [sv]
Objektdetektering, Maskininlärning, Neurala nätverk, Artificiell intelligens, Python
National Category
Engineering and Technology
Identifiers
URN: urn:nbn:se:hb:diva-23576OAI: oai:DiVA.org:hb-23576DiVA, id: diva2:1453151
Supervisors
Examiners
Available from: 2020-07-09 Created: 2020-07-09 Last updated: 2020-07-09Bibliographically approved

Open Access in DiVA

No full text in DiVA

By organisation
Faculty of Textiles, Engineering and Business
Engineering and Technology

Search outside of DiVA

GoogleGoogle Scholar

urn-nbn

Altmetric score

urn-nbn
Total: 477 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • harvard-cite-them-right
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Other style
More styles
Language
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Other locale
More languages
Output format
  • html
  • text
  • asciidoc
  • rtf