Simultaneous in-plane motion estimation and point matching using geometric cues only

Authors:
Pierre Fite Georgel;Adrien Bartoli;Nassir Navab
Affiliations:
CAMP, TU München;Université d'Auvergne;CAMP, TU München
Venue:
WMVC'09 Proceedings of the 2009 international conference on Motion and video computing
Year:
2009

Citing 13
Cited 1

Some Properties of the E Matrix in Two-View Motion Estimation

IEEE Transactions on Pattern Analysis and Machine Intelligence
Determining the Epipolar Geometry and its Uncertainty: A Review

International Journal of Computer Vision
Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography

Communications of the ACM
Geometric Hashing: An Overview

IEEE Computational Science & Engineering
The Softassign Procrustes Matching Algorithm

IPMI '97 Proceedings of the 15th International Conference on Information Processing in Medical Imaging
In defence of the 8-point algorithm

ICCV '95 Proceedings of the Fifth International Conference on Computer Vision
A Probabilistic Criterion to Detect Rigid Point Matches Between Two Images and Estimate the Fundamental Matrix

International Journal of Computer Vision
Distinctive Image Features from Scale-Invariant Keypoints

International Journal of Computer Vision
Epipolar Geometry Estimation via RANSAC Benefits from the Oriented Epipolar Constraint

ICPR '04 Proceedings of the Pattern Recognition, 17th International Conference on (ICPR'04) Volume 1 - Volume 01
A Comparison of Affine Region Detectors

International Journal of Computer Vision
A tutorial on spectral clustering

Statistics and Computing
Speeded-Up Robust Features (SURF)

Computer Vision and Image Understanding
Pose Priors for Simultaneously Solving Alignment and Correspondence

ECCV '08 Proceedings of the 10th European Conference on Computer Vision: Part II

Simultaneous Camera Pose and Correspondence Estimation with Motion Coherence

International Journal of Computer Vision

Quantified Score

Hi-index	0.00

Visualization

Abstract

We present a novel approach that, given two sets of unmatched keypoints, simultaneously estimates the in-plane camera motion and keypoint matches without using photometric information. Standard approaches estimate the epipolar geometry based on putative matches, first established with photometric information, then accepted or rejected using the epipolar constraint. Our method discretizes the space of essential matrices at different levels. It searches for the essential matrix and key-point matches which are the most geometrically coherent. We maximize geometric coherence, that we define as the number of points that can be matched based on the epipolar and unicity constraints. We applied this general framework to sets of images acquired by a moving tripod. We present promising results on simulated and real data.