Affine Covariant Features

Cookies

This website uses Google Analytics to help us improve the website content. This requires the use of standard Google Analytics cookies, as well as a cookie to record your response to this confirmation request. If this is OK with you, please click 'Accept cookies', otherwise you will see this notice on every page. For more information, please click here.

Affine Covariant Features

University of Oxford Katholieke Universiteit Leuven Inria Rhone-Alpes Czech Technical University
Collaborative work between: the Visual Geometry Group, Katholieke Universiteit Leuven, Inria Rhone-Alpes and the Center for Machine Perception.

Overview

This page is focused on the problem of detecting affine invariant features in arbitrary images and on the performance evaluation of region detectors/descriptors.

Affine Covariant Regions

affine regions from
veiwpoint 1
Image 1

affine regions from
viewpoint 2
Image 2

Publications

Region detectors

Region descriptors

  • SIFT: D. Lowe, Distinctive image features from scale invariant keypoints. In IJCV 60(2):91-110, 2004. PDF

Performance evaluation

Software

Test Data

The packages contain images in PPM format and homographies between image pairs. Data description.

Blur


bikes
1000x700
6 images

Blur


trees
1000x700
6 images

Viewpoint


graffiti
800x640
6 images

Viewpoint


bricks
1000x700
6 images

Zoom+rotation


bark
765x512
6 images

Zoom+rotation


boat
800x640
6 images

Light


cars
921x614
6 images

JPEG compression


ubc
800x640
6 images
Page maintained by km@robots.ox.ac.uk.
Last updated 15th July 2007