Research And Application On Mixture Feature Based Multi-view Remote Sensing Image Registration Algorithm

Image registration is the process of matching two images obtained from different perspectives or?and?different time,or?and?different sensors in the same scene.At present,image registration technology is mainly used for pattern recognition,computer vision,medical and military image processing and geographic information system.Remote sensing image registration as a key technology in remote sensing image processing is being widely used in different fields.Such as remote sensing image fusion,remote sensing image matching,object recognition,environmental monitoring,topographic change detection,map updating and missile guidance,etc.However,the ground relief variations and imaging viewpoint changes can result in the problems of non-rigid distortion and low overlap between remote sensing images,which further increases the difficulty of remote sensing image registration.To address the above problem,we propose a remote sensing image registration based on geometric structure feature and intensity information which contains the following contributions.?i?A multiple features based finite mixture model is constructed for dealing with different types of image features.?ii?Two different types features are combined and substituted into the mixture model to form a feature complementation,i.e.,the Euclidean distance and shape context are used to measure the similarity of geometric structure,and the SURF?Speed-up robust features?distance which is endowed with the intensity information is used to measure the scale space extrema.?iii?To prevent the ill-posed problem,a geometric constraint term is introduced into theL₂E-based energy function for better behaving the non-rigid transformation.We evaluated the performances of the proposed method by two series of remote sensing images obtained from the unmanned aerial vehicle?UAV?and Google Earth,and compared with five state-of-the-art methods?SIFT,SURF,CPD,RSOC,GLMDTPS?where our method not only improves the matching precision and in most cases all showed the best effect of registration.