Sampling minimal subsets with large spans for robust estimation

Simple item page
dc.contributor.authorQuoc, H.
dc.contributor.authorChin, T.
dc.contributor.authorChojnacki, W.
dc.contributor.authorSuter, D.
dc.date.issued2014
dc.description.abstractWhen sampling minimal subsets for robust parameter estimation, it is commonly known that obtaining an all-inlier minimal subset is not sufficient; the points therein should also have a large spatial extent. This paper investigates a theoretical basis behind this principle, based on a little known result which expresses the least squares regression as a weighted linear combination of all possible minimal subset estimates. It turns out that the weight of a minimal subset estimate is directly related to the span of the associated points. We then derive an analogous result for total least squares which, unlike ordinary least squares, corrects for errors in both dependent and independent variables. We establish the relevance of our result to computer vision by relating total least squares to geometric estimation techniques. As practical contributions, we elaborate why naive distance-based sampling fails as a strategy to maximise the span of all-inlier minimal subsets produced. In addition we propose a novel method which, unlike previous methods, can consciously target all-inlier minimal subsets with large spans.
dc.description.statementofresponsibilityQuoc Huy Tran, Tat-Jun Chin, Wojciech Chojnacki, David Suter
dc.identifier.citationInternational Journal of Computer Vision, 2014; 106(1):93-112
dc.identifier.doi10.1007/s11263-013-0643-y
dc.identifier.issn0920-5691
dc.identifier.issn1573-1405
dc.identifier.orcidChojnacki, W. [0000-0001-7782-1956]
dc.identifier.orcidSuter, D. [0000-0001-6306-3023]
dc.identifier.urihttp://hdl.handle.net/2440/88762
dc.language.isoen
dc.publisherSpringer Verlag
dc.rights© Springer Science+Business Media New York 2013
dc.source.urihttps://doi.org/10.1007/s11263-013-0643-y
dc.subjectLeast squares
dc.subjectTotal least squares
dc.subjectMinimal subsets
dc.subjectRobust fitting
dc.subjectHypothesis sampling
dc.titleSampling minimal subsets with large spans for robust estimation
dc.typeJournal article
pubs.publication-statusPublished

Files

Original bundle
Now showing 1 - 1 of 1
No Thumbnail Available
Name:
RA_hdl_88762.pdf
Size:
1.75 MB
Format:
Adobe Portable Document Format
Description:
Restricted Access
Download

Collections

Computer Science publications