umu.sePublications
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
Calibration of a Continuous Wave Time-of-Flight Camera for Motion Capturing
Umeå University, Faculty of Science and Technology, Department of Physics.
2019 (English)Independent thesis Advanced level (professional degree), 20 credits / 30 HE creditsStudent thesis
Abstract [en]

The study of human motion requires an accurate 3D positional estimate of the movement. A modern method is to utilize a continuous wave (CW) time-of-flight (ToF) camera. However, various intrinsic hardware features, or systematic errors, negatively affects the depth estimation of ToF cameras, in which, two large contributors are lens distortion and wiggling. Effects of lens distortion occurs when irregular shapes of the optical components are not accounted for, which causes deviations in the depth estimation of a ToF camera. The wiggling error arises due to an imperfect signal, generated by the complementary metal oxide semiconductor within a ToF camera’s light emitting diode. In this thesis, the aim is to obtain a reliable depth estimation of a CW ToF camera, by reducing the effects of systematic errors such as lens distortion and wiggling. To do so, lens distortion is modelled using the Brown-Conrady barrel distortion model to second order. Basis splines are used to model the camera’s depth estimation, which can be used to reduce the wiggling by constructing a lookup table based on its inflicted offsets. To test the intrinsic calibrations, the distortion caused by a stationary object was studied before and after the Brown-Conrady model was applied. Results showed that the standard deviation of a horizontal cross-section and a vertical cross-section were reduced with 31% and 43%, respectively. Additionally, a constant velocity object was analyzed to review the impact of the wiggling calibration. The standard deviation of the calibrated velocity was reduced by 78% compared to the uncalibrated velocity. Furthermore, when tracking human motion, the measurements performed by the ToF camera can be misleading depending on the camera’s perspective relative to the motion. Therefore, an extrinsic calibration method was developed, ”Principal direction rotation” (PDR), which is based on the theoretical principles behind Principal component analysis (PCA). To test this, a constant velocity movement was recorded in different camera angles, where the purpose was to achieve similar mean velocities for all angles after PDR was applied. Testing the extrinsic calibration resulted in less than 0.6% difference in mean velocity between the different camera angles. To validate the performance of the camera’s calibrated depth estimation, an additional reference device was used. A short distance sprint was recorded by both measurement techniques and their respective maximum velocities were compared by performing a minimal residual exponential fit of each velocity profile. The ToF camera measured a maximum velocity of 7.27m/s, whereas the reference device measured a maximum velocity of 7.24m/s. The accuracy of the ToF camera’s depth estimation can be improved through calibration of systematic errors, and these results demonstrate that the camera can be used to study human motion accurately.

Place, publisher, year, edition, pages
2019. , p. 44
National Category
Engineering and Technology
Identifiers
URN: urn:nbn:se:umu:diva-157148OAI: oai:DiVA.org:umu-157148DiVA, id: diva2:1295767
External cooperation
Photon Sports Technologies AB
Subject / course
Examensarbete i teknisk fysik
Educational program
Master of Science Programme in Engineering Physics
Supervisors
Examiners
Available from: 2019-03-13 Created: 2019-03-12 Last updated: 2019-03-13Bibliographically approved

Open Access in DiVA

No full text in DiVA

Search in DiVA

By author/editor
Lundström, Viktor
By organisation
Department of Physics
Engineering and Technology

Search outside of DiVA

GoogleGoogle Scholar

urn-nbn

Altmetric score

urn-nbn
Total: 46 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