Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • 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
Artificial intelligence application for objects identification
KTH, School of Engineering Sciences (SCI), Aeronautical and Vehicle Engineering, Vehicle Dynamics.
2016 (English)Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesis
Abstract [en]

New Cities expand rapidly and the increase of population is resulting in longer and more frequent travelling distances. The demand of increasing the vehicles’ efficiency, safety and comfort is highly prioritized. Therefore, the demand of au-tonomous vehicles is increasing, since many studies indicate of reducing road acci-dent by eliminating the human errors. This thesis is mainly about a study of Con-volution neural network and how it is used for the purpose of object identification. Also, a customized model has been created used for objects identification found in the Swedish traffic, which are cars, people, traffic lights and Swedish traffic signs. The choice of the network has been decided based on specific criteria, which are; the network should not be large and is still sufficient enough in terms of accuracy and processing speed. In this thesis, based on the results analysis, the criteria are partly fulfilled and in the discussion section, the drawbacks are presented and what should be done before the model can be implemented on an embedded system.

Abstract [sv]

Nya städer expanderar snabbt och ökningen av befolkningen resulterar i längre och mer frekventa resor. Kravet på att öka fordonens effektivitet, säkerhet och kom-fort har därför hög prioritet. Detta skapar i sin tur en stor efterfrågan på autonoma fordon, eftersom många studier tyder på att reducering av trafikolyckor är möjlig genom eliminering av mänskliga fel. Denna avhandling handlar huvudsakligen om en studie av artificiella neurala nätverk och hur de används för att identifiera objekt. Dessutom har en anpassad modell skapats som används för identifiering av objekt som finns i den svenska trafiken, dessa objekt är; bilar, personer, trafikljus och svenska trafikskyltar. Valet av nätverket har bestämts utifrån specifika kriterier. Nätverket ska inte vara stort men fortfarande tillräckligt bra när det gäller noggrannhet och hastighet av detektering. Baserad på resultatanalysen, uppfylls kriterierna delvis och i diskussionen presenteras svagheterna och vad som ska åtgärdas innan mod-ellen kan implementeras på ett inbyggt system.

Place, publisher, year, edition, pages
2016. , p. 56
Series
TRITA-SCI-GRU ; 2019:344
National Category
Vehicle Engineering
Identifiers
URN: urn:nbn:se:kth:diva-265623OAI: oai:DiVA.org:kth-265623DiVA, id: diva2:1380216
External cooperation
Alten Sverige AB
Examiners
Available from: 2019-12-19 Created: 2019-12-18 Last updated: 2019-12-19Bibliographically approved

Open Access in DiVA

fulltext(15297 kB)45 downloads
File information
File name FULLTEXT01.pdfFile size 15297 kBChecksum SHA-512
8010ba8484a779fa53b7d06a0d36642a071aa02290f876d990ecbf9bcffd35eabaf2972eb67b10eee1b1ae9d0d4c61953627fd36c91b17d6a5c1a01b4f5f1efd
Type fulltextMimetype application/pdf

By organisation
Vehicle Dynamics
Vehicle Engineering

Search outside of DiVA

GoogleGoogle Scholar
Total: 45 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

urn-nbn

Altmetric score

urn-nbn
Total: 75 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • 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