Object Pose Estimation using Variants of Patch-Duplet and SIFT Descriptors
2010 (English)Report (Other academic)
Recent years have seen a lot of work on local descriptors. In all published comparisons or evaluations, the now quite well-known SIFT-descriptor has been one of the top performers. For the application of object pose estimation, one comparison showed a local descriptor, called the Patch-duplet, of equal or better performance than SIFT. This paper examines different properties of those two descriptors by constructing and evaluating hybrids of them. We also extend upon the object pose estimation experiments of the original Patch-duplet paper. All tests use real images. We also show what impact camera calibration and image rectification has on an application such as object pose estimation. A new feature based on the Patch-duplet descriptor and the DoG detector emerges as the feature of choice under illuminiation changes in a real world application.
Place, publisher, year, edition, pages
2010. , 15 p.
LiTH-ISY-R, ISSN 1400-3902 ; 2950
Computer Vision and Robotics (Autonomous Systems)
IdentifiersURN: urn:nbn:se:liu:diva-57331ISRN: LiTH-ISY-R-2950OAI: oai:DiVA.org:liu-57331DiVA: diva2:325005
This is an extension of work found in http://urn.kb.se/resolve?urn=urn:nbn:se:liu:diva-562682010-06-162010-06-162010-06-23