Finding the Epipoles of a Sequence of Images for Forward Moving Camera

Full Text PDF PDF
Author(s) MD. Hamidur Rahman | Sarat C. Uppolapati | Ravitez R. Gopavaram | Wlodek J. Kulesza
Pages 787-790
Volume 2
Issue 5
Date May, 2012
Keywords Camera Motion, Epipolar Line, Epipole, Fundamental Matrix, SIFT


Finding epipoles of a sequence of images has been an important topic of considerable research in the field of computer vision. Epipole is the main feature in epipolar geometry which has enormous demand in computer vision. In most of the cases epipole is estimated from two images and the images are taken by two stationary cameras or by one stationary camera from two different locations. But if the images are taken from a moving camera, the constraints of images are changed and finding epipoles becomes very difficult. A better performance can be achieved by capturing a sequence of images for a small translation of the camera in pure forward direction then the epipole can be found for each image. The implementation of our technique deals with the estimation of epipoles for a sequence of images for the pure forward moving camera.

< Back to May Issue