Change search
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • harvard1
  • 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
Monocular Image Anthropometry: Evaluating the estimation of human body measurements using images of subjects in canonical poses taken with calibrated RGB cameras
KTH, School of Computer Science and Communication (CSC).
2017 (English)Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesisAlternative title
Monokulär bildbaserad antropometri (Swedish)
Abstract [en]

Estimating anthropometric measurements from images is an important task with multiple commercial applications (e.g. sizing recommendations of garments in online shopping). This thesis evaluates a novel silhouette based method for extracting tailor measurements (e.g. chest circumference or inseam) from images of subjects in canonical poses taken with a single, calibrated RGB camera. The measurement system is optimized for images taken with the camera on a smartphone which is sitting on the floor, tilted against a wall.

Using this method, anthropometric measurements of human- and humanoid subjects are estimated with a mean absolute error of approximately 1.4 cm, making this method a viable alternative to measurement by a (novice) human measurer using a measuring tape.

Abstract [sv]

Uppskattningen av kroppsmått utifrån bilder har många kommersiella tillämpningar som t.ex. storleksrekommendationer för kläder. Här evalueras en ny metod för att utvinna antropometriska mätningar från bilder som tagits med en kalibrerad RGM-kamera. Metoden uppskattar kroppsmåttmed 1.4 cm felmarginal (mean absolute error) och kan därför användas istället för måttband i vissa situationer.

Place, publisher, year, edition, pages
2017. , p. 39
National Category
Computer Sciences
Identifiers
URN: urn:nbn:se:kth:diva-217285OAI: oai:DiVA.org:kth-217285DiVA, id: diva2:1155018
External cooperation
Mesher ehf.
Educational program
Master of Science - Machine Learning
Supervisors
Examiners
Available from: 2017-11-09 Created: 2017-11-06 Last updated: 2018-01-13Bibliographically approved

Open Access in DiVA

No full text in DiVA

By organisation
School of Computer Science and Communication (CSC)
Computer Sciences

Search outside of DiVA

GoogleGoogle Scholar

urn-nbn

Altmetric score

urn-nbn
Total: 18 hits
CiteExportLink to record
Permanent link

Direct link
Cite
Citation style
  • apa
  • harvard1
  • 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