A CONTOUR-BASED APPROACH TO MULTISENSOR IMAGE REGISTRATION

Image registration is concerned with the establishment of corresponden ce between images of the same scene, One challenging problem in this a rea is the registration of multispectral/multisensor images. In genera l, such images have different gray level characteristics, and simple t echniques such as those based on area correlations cannot be applied d irectly, On the other hand, contours representing region boundaries ar e preserved in most cases, In this paper, we present two contour based methods which use region boundaries and other strong edges as matchin g primitives, The first contour matching algorithm is based on the cha in-code correlation and other shape similarity criteria such as invari ant moments, Closed contours and the salient segments along the open c ontours are matched separately, This method works well for image pairs in which the contour information is well preserved, such as the optic al images from Landsat and Spot satellites. For the registration of th e optical images with synthetic aperture radar (SAR) images, we propos e an elastic contour matching scheme based on the active contour model . Using the contours from the optical image as the initial condition, accurate contour locations in the SAR image are obtained by applying t he active contour model, Both contour matching methods are automatic a nd computationally quite efficient, Experimental results with various kinds of image data have verified the robustness of our algorithms, wh ich have outperformed manual registration in terms of root mean square error at the control points.