Estimating the fundamental matrix using L∞ minimization algorithm

Fundamental matrix estimation is a central problem in computer vision and forms the basis of tasks such as stereo imaging and structure from motion. A new method for the estimation of the fundamental matrix from point correspondences is presented. The minimization of an objective function closer to the geometric distance is performed based L_infin minimization framework. The fundamental matrix is optimally computed with taking into account the rank-two constraint, and the method is no need for normalization of the image coordinates. It is shown how this nonlinearly estimating the fundamental matrix can be solved avoiding local minima by using semidefinite programming. Experiments on real images show that this method provides a more accurate estimate of the fundamental matrix and superior to previous approaches.