FMPR-Net: False Matching Point Removal Network for Very-High-Resolution Satellite Image Registration

Image registration is the most basic preprocessing method used to unify coordinates among multitemporal very-high-resolution (VHR) satellite images, thus allowing the acquisition of reliable data on the Earth's surface. Although image registration requires multiple matching points (MPs), false MPs (FMPs) are included because of the similar spectral patterns and noise. However, removing FMPs from VHR satellite image pairs is challenging, especially when the images are directly affected by complex factors, such as shadow, relief displacement, and terrain shielding. Therefore, we propose an FMP removal network (FMPR-net) based on deep learning to eliminate effectively the FMPs to improve registration accuracy. The training dataset is produced by a semiautomatic method. It involves the generation of image patch pairs based on a matching process of scale-invariant feature transform (SIFT) and the assignment of labels referring to the characteristics of true MPs (TMPs) and FMPs. The FMPR-net is designed in a Siamese format consisting of two matching point deep feature extractors (MDFEs). The architecture of the MDFE consists of one main network and three branch networks to achieve robust extraction of meaningful deep features describing the characteristics of MPs. The FMPR-net removes the FMPs using a true matching probability calculated based on the similarity between deep features. Experiments conducted on four pairs of VHR satellite images have demonstrated that the FMPR-net can effectively remove the FMPs. Consequently, accurate VHR satellite image registration is possible by reducing uncertainty caused by the FMPs.