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
A novel shape matching descriptor for real-time hand gesture recognition
KTH, School of Electrical Engineering and Computer Science (EECS).
2018 (English)Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesis
Abstract [en]

In object recognition applications in computer vision, feature information extractedfrom images is used to dierentiate between objects. One of the mostimportant features that can be extracted from an object is its shape. ShapeMatching is a subset of object recognition methodology and is an area that hasattracted a lot of research interest over the years. Shape matching has showngreat potential in scenarios where images lack texture and colour informationor cannot be trained by learning models of feature combinations. However, themost accurate shape matching algorithms found in literature, such as matchingwith Shape Context, tend to be computationally inecient having polynomialcomputational complexities. This thesis addresses the problem of real-timehand gesture recognition and investigates whether shape matching can providea robust and computationally ecient method for real-time hand gesture recognition.Several shape matching algorithms were investigated and a novel shapematching algorithm is proposed as the optimal solution for this problem.

Abstract [sv]

I objektigenkänningsapplikationer i datorvision används funktioner som extra-heras från bilder för att skilja mellan objekt. En av de viktigaste egenskaperna som kan extraheras från ett objekt är dess form. Shape Matching är en delmängd av objektigenkänningsmetodik och är ett omrde som har lockat mycket forskningsintresse genom åren. Form matchning har visat stor potential i scenarier där bilder saknar textur och färginformation eller kan inte utbildas genom att lära sig modeller av funktionskombinationer. De mest exakta form matchningsalgoritmerna som finns i litteraturen, till exempel matchning med Shape Context, tenderar att vara beräkningsmässigt ineffektiva med polynomala berkningskomplexiteter. Denna avhandling tar upp problemet med realtidsigenkänning i realtid och undersöker om formmatchning kan ge en robust och beräkningsmssig effektiv metod för realtidsigenkänning av rektangel. Flera former för matchande algoritmer undersöktes och en ny form matchande algoritm föreslås som denoptimala lösningen för detta problem.

Place, publisher, year, edition, pages
2018. , p. 68
Series
TRITA-EECS-EX ; 2018:175
National Category
Engineering and Technology
Identifiers
URN: urn:nbn:se:kth:diva-253265OAI: oai:DiVA.org:kth-253265DiVA, id: diva2:1324323
External cooperation
Manomotion
Examiners
Available from: 2019-06-13 Created: 2019-06-13 Last updated: 2019-06-13Bibliographically approved

Open Access in DiVA

fulltext(3725 kB)22 downloads
File information
File name FULLTEXT01.pdfFile size 3725 kBChecksum SHA-512
3ad04434b1d4a324ed9d0ab2be711f3e0ec5aa564e322d27b8319a2790db727310f5f82bdb21d751ea2a8d74c60a67cc52fb2f7f1e66fe794097fc85973d8054
Type fulltextMimetype application/pdf

By organisation
School of Electrical Engineering and Computer Science (EECS)
Engineering and Technology

Search outside of DiVA

GoogleGoogle Scholar
Total: 22 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: 36 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