Fourier-based image correlation for image registration

Abstract

Image registration is an important technique in many computer vision applications, such as image fusion, object tracking, face recognition, change detection, etc. Registration of multi-date images is based on primitive space, similarity measure, search and optimization strategy. Each component plays a fundamental role in estimating the best spatial transformation, which has a direct impact on the robustness and accuracy of these methods. In this paper, we will be discussing classical and recent image registration methods, including their fundamental principles. This review provides a comprehensive reference resource for researchers involved in image registration with Fourier-based image correlation by describing Fourier-based image correlation methods, describing existing subpixel techniques in the frequency domain, and summarizing comparative studies of subpixel techniques.

Keywords: sub-pixel registration, matching, phase correlation, Fourier transform