Research Of Image Registration Method Based On Edge Feature

Image registration is a fundamental task in image processing,which is used to match two or multiple more images, for example, at different times, from different sensors or from different view points. The diversity of detected features is a main problem, and the features extraction should be of robustness and accuracy for approach to image registration based on features.Edge detection is widely applied to image registration,which is very important to the matched accuracy of images. Edge detection is deeply studied in this paper, and we place an emphasis on contrasting modern methods of edge detection. Cellular Automata edge detection method is improved, and wavelet transform edge detection is selected to image matching. On the basis of practical alignment metric,a new approach to image registration based on combination of wavelet analysis and alignment metric is proposed,and then infrared/visual image registration is attempted to have an explorative exppriment.Several experiments conducted illustrate that this method has a good performance of efficiency,accuracy,and it can also be applied to many situations.