Image Registration on Satellite Images

Image registration is a key operation to spatially align two or more images for comparing the difference between them or exploiting complementary information from those images. The main objective regarding automatic registration of satellite images is to obtain an accurate set of tie points and then apply the transformation function which is most suitable to the pair of images to be registered. A considerably large number of approaches may be found in the literature regarding the automatic obtention of tie points being mainly areaor feature-based methods by means of the image intensity values in their close neighborhoods, the feature spatial distribution, or the feature symbolic description. Automatic image registration is still a present challenge for the remote sensing community. Although a wide variety of image registration methods have been proposed in the last few years, there are several drawbacks which avoid their common use in practice. The recently proposed scale invariant feature transform (SIFT) approach has already revealed to be a powerful tool for retrieval of tie points in general image processing tasks, but it has a limited performance when directly applied to remote sensing images. A robust and efficient method for AIR(Automatic Image Registration) is proposed, which combines image segmentation and SIFT, complemented by an outlier removal stage. The reference and unregistered images may differ in translation, rotation, and scale and may present distortions. The performance of this method is evaluated through measures such as Nred, RMSall, RMSLOO, Pquad, Bad Point Proportion: BPP(r), Scat, Cost Function (φ).