Image Matching Navigation Guidance

Image registration based navigation composes with other navigation methods has higher precision and reliability. This dissertation is dedicated to the algorithm research of image registration applied to navigation. Information such as attitude angle of aircraft is used to emendate the large-geometry-distorted image, so the research of this dissertation is based on the hypothesis that images used have small geometry distortion.The main body of the algorithm is the one based on wavelet multi-scale analysis, which obtained good performance in the conditions of noise added, small geometry distortion and significant gray distortion. On the base of wavelet multi-scale analysis based image registration method, also the other techniques like invariant moment, image selection and genetic algorithm are studied. That GA deceptive problem exists when GAs is applied to image registration, and invariant moment is invalid when image gray distortion exists are clearly indicated. Finally a set of tailorable algorithms is given, which centers on the wavelet multi-scale analysis and integrates other registration techniques.