Automatic registration of high-resolution images in urban areas using local properties of features

We propose an automatic image-to-image registration of high-resolution satellite images using local properties and geometrical locations of matching points to improve the registration accuracy. First, coefficients of global affine transformation between images are extracted using a scale-invariant feature transform (SIFT)-based method, and features of the sensed image are transformed to the reference coordinate system using these coefficients. Then, a spatial distance between a feature of the reference and features of the sensed images that have been transformed to the reference coordinates within a predefined buffer is additionally used to extract precise matching points. Finally, the spatial distance integrated with Euclidean distances of invariant vectors is employed for local matching. The optimal ranges of the proposed distance and the radius of the buffer for local matching are determined using a registration consistency measure. The average orientation differences between matching points of the two images are used for outlier elimination. A mapping function model consisting of an affine transformation and piecewise linear functions is applied to the matching points for automatic registration of high-resolution images. The proposed method can extract precise matching points and gives better registration results than the SIFT-based method alone.