Image Registration with Hyperspectral Data Based on Fourier-Mellin Transform

Hyperspectral imagery is playing more and more important role in many fields such as geology, agriculture, environment, military, atmosphere and so on. We need to register Hyperspectral imagery so that we can build coherent image cubes and get coherent spectrum of pixels. Our concern in this paper is the Fourier-Mellin Transform (FMT) for image registration process and its related techniques for real hyperspectral data. FMT can recover translation, rotation. It is computationally efficient. Techniques such as filter, window and downscale for improving precision in the registration process are taken/put forward and discussed. With highpass emphasis filter and downscale for registration of real hyperspectral data can get a right or better result in some situation.