A new framework for constructing accurate affine invariant regions

Research output: Contribution to journalArticle

Abstract

In this study, we propose a simple, yet general and powerful framework for constructing accurate affine invariant regions. In our framework, a method for extracting reliable seed points is first proposed. Then, regions which are invariant to most common affine transformations can be extracted from seed points by two new methods the Path Growing (PG) or the Thresholding Seeded Growing Region (TSGR). After that, an improved ellipse fitting method based on the Direct Least Square Fitting (DLSF) is used to fit the irregularly-shaped contours from the PG or the TSGR to obtain ellipse regions as the final invariant regions. In the experiments, our framework is first evaluated by the criterions of Mikolajczyk's evaluation framework [1], and then by near-duplicate detection problem [2]. Our framework shows its superiorities to the other detectors for different transformed images under Mikolajczyk's evaluation framework and the one with TSGR also gives satisfying results in the application to near-duplicate detection problem.

Original languageEnglish
Pages (from-to)1831-1840
Number of pages10
JournalIEICE Transactions on Information and Systems
VolumeE90-D
Issue number11
DOIs
Publication statusPublished - 2007 Nov

Fingerprint

Seed
Detectors
Experiments

Keywords

  • Affine invariant region
  • Ellipse fitting
  • Near-duplicate detection
  • Path Growing
  • Thresholding seeded growing regions

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Software
  • Artificial Intelligence
  • Hardware and Architecture
  • Computer Vision and Pattern Recognition

Cite this

A new framework for constructing accurate affine invariant regions. / Tian, Li; Kamata, Seiichiro.

In: IEICE Transactions on Information and Systems, Vol. E90-D, No. 11, 11.2007, p. 1831-1840.

Research output: Contribution to journalArticle

@article{335ef68c4f74488a98fd55541aee6eeb,
title = "A new framework for constructing accurate affine invariant regions",
abstract = "In this study, we propose a simple, yet general and powerful framework for constructing accurate affine invariant regions. In our framework, a method for extracting reliable seed points is first proposed. Then, regions which are invariant to most common affine transformations can be extracted from seed points by two new methods the Path Growing (PG) or the Thresholding Seeded Growing Region (TSGR). After that, an improved ellipse fitting method based on the Direct Least Square Fitting (DLSF) is used to fit the irregularly-shaped contours from the PG or the TSGR to obtain ellipse regions as the final invariant regions. In the experiments, our framework is first evaluated by the criterions of Mikolajczyk's evaluation framework [1], and then by near-duplicate detection problem [2]. Our framework shows its superiorities to the other detectors for different transformed images under Mikolajczyk's evaluation framework and the one with TSGR also gives satisfying results in the application to near-duplicate detection problem.",
keywords = "Affine invariant region, Ellipse fitting, Near-duplicate detection, Path Growing, Thresholding seeded growing regions",
author = "Li Tian and Seiichiro Kamata",
year = "2007",
month = "11",
doi = "10.1093/ietisy/e90-d.11.1831",
language = "English",
volume = "E90-D",
pages = "1831--1840",
journal = "IEICE Transactions on Information and Systems",
issn = "0916-8532",
publisher = "Maruzen Co., Ltd/Maruzen Kabushikikaisha",
number = "11",

}

TY - JOUR

T1 - A new framework for constructing accurate affine invariant regions

AU - Tian, Li

AU - Kamata, Seiichiro

PY - 2007/11

Y1 - 2007/11

N2 - In this study, we propose a simple, yet general and powerful framework for constructing accurate affine invariant regions. In our framework, a method for extracting reliable seed points is first proposed. Then, regions which are invariant to most common affine transformations can be extracted from seed points by two new methods the Path Growing (PG) or the Thresholding Seeded Growing Region (TSGR). After that, an improved ellipse fitting method based on the Direct Least Square Fitting (DLSF) is used to fit the irregularly-shaped contours from the PG or the TSGR to obtain ellipse regions as the final invariant regions. In the experiments, our framework is first evaluated by the criterions of Mikolajczyk's evaluation framework [1], and then by near-duplicate detection problem [2]. Our framework shows its superiorities to the other detectors for different transformed images under Mikolajczyk's evaluation framework and the one with TSGR also gives satisfying results in the application to near-duplicate detection problem.

AB - In this study, we propose a simple, yet general and powerful framework for constructing accurate affine invariant regions. In our framework, a method for extracting reliable seed points is first proposed. Then, regions which are invariant to most common affine transformations can be extracted from seed points by two new methods the Path Growing (PG) or the Thresholding Seeded Growing Region (TSGR). After that, an improved ellipse fitting method based on the Direct Least Square Fitting (DLSF) is used to fit the irregularly-shaped contours from the PG or the TSGR to obtain ellipse regions as the final invariant regions. In the experiments, our framework is first evaluated by the criterions of Mikolajczyk's evaluation framework [1], and then by near-duplicate detection problem [2]. Our framework shows its superiorities to the other detectors for different transformed images under Mikolajczyk's evaluation framework and the one with TSGR also gives satisfying results in the application to near-duplicate detection problem.

KW - Affine invariant region

KW - Ellipse fitting

KW - Near-duplicate detection

KW - Path Growing

KW - Thresholding seeded growing regions

UR - http://www.scopus.com/inward/record.url?scp=68249156901&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=68249156901&partnerID=8YFLogxK

U2 - 10.1093/ietisy/e90-d.11.1831

DO - 10.1093/ietisy/e90-d.11.1831

M3 - Article

AN - SCOPUS:68249156901

VL - E90-D

SP - 1831

EP - 1840

JO - IEICE Transactions on Information and Systems

JF - IEICE Transactions on Information and Systems

SN - 0916-8532

IS - 11

ER -