Geometrically Based Linear Iterative Clustering for Quantitative Feature Correspondence

dc.contributor.authorYan, Qinganen_US
dc.contributor.authorYang, Longen_US
dc.contributor.authorLiang, Chaoen_US
dc.contributor.authorLiu, Huajunen_US
dc.contributor.authorHu, Ruiminen_US
dc.contributor.authorXiao, Chunxiaen_US
dc.contributor.editorEitan Grinspun and Bernd Bickel and Yoshinori Dobashien_US
dc.date.accessioned2016-10-11T05:18:35Z
dc.date.available2016-10-11T05:18:35Z
dc.date.issued2016
dc.description.abstractA major challenge in feature matching is the lack of objective criteria to determine corresponding points. Recent methods find match candidates first by exploring the proximity in descriptor space, and then rely on a ratio-test strategy to determine final correspondences. However, these measurements are heuristic and subjectively excludes massive true positive correspondences that should be matched. In this paper, we propose a novel feature matching algorithm for image collections, which is capable of providing quantitative depiction to the plausibility of feature matches. We achieve this by exploring the epipolar consistency between feature points and their potential correspondences, and reformulate feature matching as an optimization problem in which the overall geometric inconsistency across the entire image set ought to be minimized. We derive the solution of the optimization problem in a simple linear iterative manner, where a k-means-type approach is designed to automatically generate consistent feature clusters. Experiments show that our method produces precise correspondences on a variety of image sets and retrieves many matches that are subjectively rejected by recent methods. We also demonstrate the usefulness of the framework in structure from motion task for denser point cloud reconstruction.en_US
dc.description.number7
dc.description.sectionheadersMatching and Interpolation
dc.description.seriesinformationComputer Graphics Forum
dc.description.volume35
dc.identifier.doi10.1111/cgf.12998
dc.identifier.issn1467-8659
dc.identifier.pages1-10
dc.identifier.urihttps://doi.org/10.1111/cgf.12998
dc.identifier.urihttps://diglib.eg.org:443/handle/10.1111/cgf12998
dc.publisherThe Eurographics Association and John Wiley & Sons Ltd.en_US
dc.subjectI.3.m [Computer Graphics]
dc.subjectComputational photography
dc.subject
dc.subject
dc.subjectI.4.0 [Computer Graphics]
dc.subjectGeneral
dc.subject
dc.subject
dc.subjectI.4.7 [Computer Graphics]
dc.subjectFeature Measurement
dc.titleGeometrically Based Linear Iterative Clustering for Quantitative Feature Correspondenceen_US
Files
Collections