Multi-source Remote Sensing Image Registration Based On Phase Consistency

Multi-source remote sensing image registration is a vital research subject in the field of image processing.Its main objective is to align two or more images acquired by different sensors at different time or perspectives for the same geographical object.With the development of modern technology,the technique of multi-source remote sensing image data acquisition is also improved.The accurate registration of multi-source remote sensing images can not only combine the advantages of image information,but also plays a fundamental role in image fusion,target recognition and detection.Therefore,the study of multi-source image registration technology has important research significance.In order to improve the speed and accuracy of remote sensing image registration,two registration methods based on improved Harris operator and one based on phase consistency are proposed by summarizing and analyzing remote sensing image registration technology.The details are as follows:(1)In face of the problem that corner detection by traditional Harris operator is timeconsuming,an improved Harris operator method was proposed.By taking advantage of the characteristics of small gray difference in the neighborhood of non-corner points,non-corner points are pre-eliminated first,so as to quickly extract image corner features.Then,the improved Harris operator is combined with the traditional registration algorithm to effectively speed up the extraction of common features of remote sensing images,and achieve fast registration between multi-source remote sensing images.The experimental results showed that this operator can not only get uniformly distributed corner features,but also effectively improve the registration speed and accuracy of multi-source remote sensing images.(2)Addressing the problem of nonlinear radiation difference between multi-source remote sensing images,this manuscript proposed a multi-source remote sensing image registration method based on multi-scale phase consistency.The method combines the multi-scale information of the image to construct the multi-scale space of the image,and then uses the guided filtering method to enhance the details of the image to keep the high frequency information of the image.In order to solve the problem of nonlinear difference between remote sensing images,the proposed method is inspired from the phase consistency of the edge detection method,to construct multi-scale phase consistency feature map without the intensity information,then extracts corner features in the map to get robust and uniformly distributed image angular points.Finally,the normalized cross-correlation criterion(NCC)is used to get the matching point pairs between remote sensing images,then the correct matching point pairs were obtained after removing the false matching points by the RANSAC algorithm.Experimental results show that this method can not only get uniformly distributed matching point pairs,but also achieve fast registration of multi-source remote sensing images.In this manuscript,the multi-source remote sensing image registration method was validated on three pairs of SAR images acquired by Gaofen3 and Google earth optical images,a pair of optical image and laser depth image,a pair of optical image and infrared image Experiments demonstrated that the proposed registration method is tractable,effective and accurate.