RMIT University
Browse

Investigating Keypoint Repeatability for 3D Correspondence Estimation in Cluttered Scenes

conference contribution
posted on 2024-11-03, 11:25 authored by Quang Chiem, Richardt WilkinsonRichardt Wilkinson, Margaret LechMargaret Lech, Eva Cheng
In 3D object recognition, local feature-based recognition is known to be robust against occlusion and clutter. Local feature estimation requires feature correspondences, including feature extraction and matching. Feature extraction is normally a two-stage process that estimates keypoints and keypoint descriptors, and existing studies show repeatability to be a good indicator of keypoint feature detector robustness. However, the impact of keypoint repeatability on feature correspondence estimation and overall feature matching accuracy has not yet been studied. In this paper, local features are extracted at both regular and repeatable 3D keypoints using leading keypoint detectors combined with the SHOT descriptor to estimate a set of correspondences. When using a keypoint detector of high repeatability, experimental results show improved feature matching accuracy and reduced computational requirements for the feature description and matching, and overall correspondence estimation process.

History

Start page

134

End page

139

Total pages

6

Outlet

Proceedings of the 2017 International Conference on Digital Image Computing: Techniques and Applications (DICTA 2017)

Name of conference

DICTA 2017

Publisher

IEEE

Place published

United States

Start date

2017-11-29

End date

2017-12-01

Language

English

Copyright

© 2017 IEEE

Former Identifier

2006089224

Esploro creation date

2020-06-22

Fedora creation date

2019-01-31

Usage metrics

    Scholarly Works

    Exports

    RefWorks
    BibTeX
    Ref. manager
    Endnote
    DataCite
    NLM
    DC