Survey on Registration Techniques of Agency for Defense Development, Yuseong P.O.Box 35, Daejeon,
305-600, Korea
Abstract This paper aims to present an overview of recent as well as classic multi-modal image registration methods, specially about registration of visible and infrared image. Image registration is the process of overlaying images (two or more) of the same scene taken at different times, from different viewpoints, and/or by different sensors. The registration geometrically aligns two images (the reference image and sensed input images). The reviewed approaches are classified according to their nature (pixel-based and feature-based). Registration algorithms compute transformations to set correspondence between the two images. The purpose of this paper is to provide a comprehensive review of the existing literature available on registration algorithms of visible and infrared images. Moreover, we compare and analyze using real scene data the performance of Mutual information based algorithm and Fourier transform based algorithm which are representative algorithms in registration algorithms. Keywords: Image
Regestration Algorithm, Mutual information based
algorithm, +: Corresponding author: Kyungsu Kim IT Convergence Practice (INPRA), Vol. 3, No. 2, pp. 25-35, June 2015 [pdf] |