liu.seSearch for publications in DiVA
Change search
ReferencesLink to record
Permanent link

Direct link
Orientation Estimation Based on Weighted Projection onto Quadratic Polynomials
Linköping University, Department of Electrical Engineering, Computer Vision. Linköping University, The Institute of Technology.
2000 (English)In: Vision, Modeling, and Visualization 2000: proceedings, 2000, 89-96 p.Conference paper (Refereed)
Abstract [en]

Essentially all Computer Vision strategies require initial computation of orientation structure or motion estimation. Although much work has been invested in this subfield, methods have so far been very computationally demanding and/or not very robust. In this paper we present a novel method for computation of orientation tensors for signals of any dimensionality. The method is based on local weighted least squares approximations of the signal by second degree polynomials. It is shown how this can be implemented very efficiently by means of separable convolutions and that the method gives very accurate orientation estimates. We also introduce the new concept of orientation functionals, of which orientation tensors is a subclass. Finally we demonstrate the critical importance of using a proper weighting function in the local projection of the signal onto polynomials.

Place, publisher, year, edition, pages
2000. 89-96 p.
National Category
Engineering and Technology
URN: urn:nbn:se:liu:diva-21750OAI: diva2:273875
5th International Fall Workshop. Vision, Modeling, and Visualization 2000, 22-24 November 2000, Saarbrücken, Germany
Available from: 2009-10-25 Created: 2009-10-05 Last updated: 2015-02-26

Open Access in DiVA

fulltext(676 kB)256 downloads
File information
File name FULLTEXT01.pdfFile size 676 kBChecksum SHA-512
Type fulltextMimetype application/pdf

Other links

Search in DiVA

By author/editor
Farnebäck, Gunnar
By organisation
Computer VisionThe Institute of Technology
Engineering and Technology

Search outside of DiVA

GoogleGoogle Scholar
Total: 256 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

Total: 213 hits
ReferencesLink to record
Permanent link

Direct link