A Robust, Feature-based Algorithm for Aerial Image Registration

In this paper an algorithm for aerial image registration is proposed. The objective of this algorithm is to register aerial images having only partial overlap, which are also geometrically distorted due to the different sensing conditions and in addition they may be contaminated with noise, may be blurred, etc. The geometric distortions considered in the registration process are rotation, translation and scaling. The proposed algorithm consists of three main steps: feature point extraction using a feature point extractor based on scale-interaction of Mexican-hat wavelets, obtaining the correspondence between the feature points of the first (reference) and the second image based on Zernike moments of neighborhoods centered on the feature points, and estimating the transformation parameters between the first and the second images using an iterative weighted least squares algorithm. Experimental results illustrate the accuracy of image registration for images with partial overlap in the presence of additional image distortions, such as noise contamination and image blurring.