Robustness of point feature detection


by Z Song, R Klette
Abstract:
This paper evaluates 2D feature detection methods with respect to invariance and efficiency properties. The studied feature detection methods are as follows: Speeded Up Robust Features, Scale Invariant Feature Transform, Binary Robust Invariant Scalable Keypoints, Oriented Binary Robust Independent Elementary Features, Features from Accelerated Segment Test, Maximally Stable Extremal Regions, Binary Robust Independent Elementary Features, and Fast Retina Keypoint. A long video sequence of traffic scenes is used for testing these feature detection methods. A brute-force matcher and Random Sample Consensus are used in order to analyse how robust these feature detection methods are with respect to scale, rotation, blurring, or brightness changes. After identifying matches in subsequent frames, RANSAC is used for removing inconsistent matches; remaining matches are taken as correct matches. This is the essence of our proposed evaluation technique. All the experiments use a proposed repeatability measure, defined as the ratio of the numbers of correct matches, and of all keypoints. © 2013 Springer-Verlag.
Reference:
Robustness of point feature detection (Z Song, R Klette), In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), volume 8048 LNCS, 2013.
Bibtex Entry:
@inproceedings{song2013robustnessdetection,
author = "Song, Z and Klette, R",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "91--99",
title = "Robustness of point feature detection",
volume = "8048 LNCS",
year = "2013",
abstract = "This paper evaluates 2D feature detection methods with respect to invariance and efficiency properties. The studied feature detection methods are as follows: Speeded Up Robust Features, Scale Invariant Feature Transform, Binary Robust Invariant Scalable Keypoints, Oriented Binary Robust Independent Elementary Features, Features from Accelerated Segment Test, Maximally Stable Extremal Regions, Binary Robust Independent Elementary Features, and Fast Retina Keypoint. A long video sequence of traffic scenes is used for testing these feature detection methods. A brute-force matcher and Random Sample Consensus are used in order to analyse how robust these feature detection methods are with respect to scale, rotation, blurring, or brightness changes. After identifying matches in subsequent frames, RANSAC is used for removing inconsistent matches; remaining matches are taken as correct matches. This is the essence of our proposed evaluation technique. All the experiments use a proposed repeatability measure, defined as the ratio of the numbers of correct matches, and of all keypoints. © 2013 Springer-Verlag.",
doi = "10.1007/978-3-642-40246-3_12",
isbn = "9783642402456",
issn = "0302-9743",
eissn = "1611-3349",
issue = "PART 2",
language = "eng",
}