A Novel Speed-up Feature Matching Algorithm for Image Registration using SUSAN and RANSAC
Monica P. Chanchlani1, Madhuri Khambete2
1Monica P. Chanchlani, Department of Electronics and Telecommunication, Cummins College of Engineering for Women, Pune, India.
2Dr. Madhuri Khambete, Principal, Cummins College of Engineering for Women, Pune, India.
Manuscript received on May 22, 2013. | Revised Manuscript received on June 07, 2013. | Manuscript published on June 30, 2013. | PP: 396-399 | Volume-2, Issue-5, June 2013. | Retrieval Number: E1874062513/2013©BEIESP

Open Access | Ethics and Policies | Cite
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)

Abstract: A novel feature matching algorithm for image registration is proposed in this paper. The accuracy of a registration process is highly dependent on the feature detection and matching. In this paper, we use a SUSAN (Smallest Univalue Segment Assimilating Nucleus) algorithm to detect features, which is one of the most excellent methods, robust to noise and less affected by rotation. One common approach used for feature matching is correlation between feature points. But in this method much computational time is required to establish the correspondences. In this paper, we overcome this difficulty through our speed-up approach. The basic concept of our approach is to calculate the descriptor values for every feature point which are then stored. These values will finally be used for feature matching. This reduces the number of operations in feature matching step and thus speed-up in matching is obtained. After matching, RANSAC method is used to find the registration transform parameters.
Keywords: Image Registration, SUSAN, Feature Matching, Computation Time, RANSAC.