Rectification on uncalibrated epipolar stereo images and dense disparity map

Authors:
Kunio Takaya
Affiliations:
University of Saskatchewan, Saskatoon, SK, Canada
Venue:
CI '07 Proceedings of the Third IASTED International Conference on Computational Intelligence
Year:
2007

Citing 3
Cited 0

Random sample consensus: a paradigm for model fitting with applications to image analysis and automated cartography

Communications of the ACM
A direct method for stereo correspondence based on singular value decomposition

CVPR '97 Proceedings of the 1997 Conference on Computer Vision and Pattern Recognition (CVPR '97)
Multiple View Geometry in Computer Vision

Multiple View Geometry in Computer Vision

Quantified Score

Hi-index	0.00

Visualization

Abstract

This paper presents a result of the studies to generate the dense disparity map based on the epipolar geometry of stereo vision. The distance (missing depth in 2D images) to a point on the objects recorded in a pair of stereo images can be estimated from the disparity due to the parallax between two cameras. The well established theories of the epipolar geometry combined with the robust method of RANSAC to calculate the fundamental matrix was implemented. Stereo images are rectified to make epipolar lines parallel along the horizontal axis (raster lines), and to enable 1D search for matching the points of correspondence. This paper proposes a scanning type pattern matching method that can be used to generate a dense disparity map which displays a 2D distribution of pixel-by-pixel disparities. Preliminary results are presented.