A General Approach for Extracting Affine Invariant Regions

CAI Hong-ping; LEI Lin; CHEN Tao; SU Yi

您当前的位置：

首页 >

文章列表页 >

A General Approach for Extracting Affine Invariant Regions

更新时间：2025-07-16

- A General Approach for Extracting Affine Invariant Regions
- Acta Electronica Sinica Vol. 36, Issue 4, Pages: 672-678(2008)
- 作者机构：
  
  国防科学技术大学电子科学与工程学院,湖南,长沙,410073
- 作者简介：
- 基金信息：
- DOI：
  CLC： TP391.4
- Published Online：25 April 2008，
  
  Published：2008
- 稿件说明：
移动端阅览
CAI Hong-ping, LEI Lin, CHEN Tao, et al. A General Approach for Extracting Affine Invariant Regions[J]. Acta Electronica Sinica, 2008, 36(4): 672-678.
DOI：

CAI Hong-ping, LEI Lin, CHEN Tao, et al. A General Approach for Extracting Affine Invariant Regions[J]. Acta Electronica Sinica, 2008, 36(4): 672-678. DOI：

摘要

本文利用尺度-空间理论和自相关矩阵的局部形状提出了一种通用的提取仿射不变特征区域的方法.首先

在尺度-空间中对图像的归一化高斯微分求三维局部极大值获得特征点和特征尺度位置

然后在特征点的特征尺度上用自相关矩阵刻画局部的灰度变化

提取的椭圆区域即为仿射不变特征区域.在此通用方法框架下构造了Harris3D、Laplace3D、Hessian3D和Localjet43D四种仿射不变特征区域算法.实验结果表明这四种算法都具有照度、旋转和尺度不变性.用本文设计的一种仿射不变性仿真实验方法验证了算法的仿射不变性.比较四种算法发现除了Harris3D性能稍差外其他三种算法性能接近.

Abstract

Using scale-space theory and the local shape of auto-correlation matrix

a new general approach is proposed to extract affine invariant regions.First

the feature points and their characteristic scales are detected by the local maxima of normal Gaussian derivatives over scale-space;Then

the auto-correlation matrices

which are used to describe local image structure

are computed on the characteristic scales of feature points.The extracted ellipse regions are affine invariant.Four affine invariant region algorithms

namely Harris3D

Laplace3D

Hessian3D and Localjet43D

are presented using the general approach.The experimental results show the four algorithms are invariant to illumination

rotation and scale changes.The affine invariance is verified by the simulation test we designed for affine invariance.Comparing the four algorithms

we find that the performance of other three algorithms is similar except for Harris3D.

关键词

Keywords

references

Views

1509

下载量

CSCD

Alert me when the article has been cited

提交

Tools

Publicity Resources

Visual Object Tracking Algorithm Based on Adaptive Feature Selection

Continual Learning Methods and Applications in Computer Vision

Open World Object Detection Based on Causal Prompt Distillation

Neural Network Based Image Style Transfer: A Survey

End-to-End Scene Text Spotting Under Dual Domain Awareness Based on Multi-Party Synergetic Explicit Information

Related Author

WANG Cai-xia

AN Qi

ZHOU Hong-ce

LI Yi-peng

FANG Yan

Wei Yun-chao

Cong Run-min

Zuo Wang-meng

Related Institution

Jilin Province Technology Research Center of Photoelectric Detection and Intelligent Information Processing

School of Electronic Information and Engineering, Changchun University of Science and Technology

School of Computer Science and Technology, Harbin Institute of Technology

School of Control Science and Engineering, Shandong University

School of Computer Science and Technology, Beijing Jiaotong University

⁰