Robust Monocular Epipolar Flow Estimation
- 1 June 2013
- conference paper
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1862-1869
- https://doi.org/10.1109/cvpr.2013.243
Abstract
We consider the problem of computing optical flow in monocular video taken from a moving vehicle. In this setting, the vast majority of image flow is due to the vehicle's ego-motion. We propose to take advantage of this fact and estimate flow along the epipolar lines of the egomotion. Towards this goal, we derive a slanted-plane MRF model which explicitly reasons about the ordering of planes and their physical validity at junctions. Furthermore, we present a bottom-up grouping algorithm which produces over-segmentations that respect flow boundaries. We demonstrate the effectiveness of our approach in the challenging KITTI flow benchmark [11] achieving half the error of the best competing general flow algorithm and one third of the error of the best epipolar flow algorithm.Keywords
This publication has 21 references indexed in Scilit:
- Trinocular optical flow estimation for intelligent vehicle applicationsPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2012
- Norm-Product Belief Propagation: Primal-Dual Message-Passing for Approximate InferenceIEEE Transactions on Information Theory, 2010
- Secrets of optical flow estimation and their principlesPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2010
- Structure- and motion-adaptive regularization for high accuracy optic flowPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2009
- Optical flow estimation with uncertainties through dynamic MRFsPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2008
- Stereo Processing by Semiglobal Matching and Mutual InformationIeee Transactions On Pattern Analysis and Machine Intelligence, 2007
- Computing visual correspondence with occlusions using graph cutsPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2002
- In defense of the eight-point algorithmIEEE Transactions on Pattern Analysis and Machine Intelligence, 1997
- Non-parametric local transforms for computing visual correspondenceLecture Notes in Computer Science, 1994
- “Determining optical flow”: a retrospectiveArtificial Intelligence, 1993