Study On Image Registration And Super-resolution Restoration Technologies

The applications of infrared and visible sensors in the fields of medical imaging,remote sensing,on-line monitoring for power equipment,human detection and tracking system have drawn more attention.Before combing two complementary sensors,correspondences between infrared and visible images should be found.However,infrared and visible images are manifestations of two different phenomena.A texture or an edge in a visible image is often missing in the infrared image because infrared cameras measure principally reflects infrared radiations emitted by objects.Therefore,the registration problem is challenging,and a novel fast and accurate method to solve the registration is important to the application of visible and infrared image registration.SIFT,adaptive support-weight and POCS methods are discussed in this paper.The aim is to improve two problems in order to enhance the imaging effect of the infrared target and improve the accuracy of image registration,which is the basis of multi-modal image sequences combination.The main work and innovation are as follows:(1)Traditional stereo matching aims to find corresponding points that are projections of the same scene point from the reference image and the target image taken from different viewing points based on RGB values of pixels.However,in a real scene,due to lighting geometry,illuminant color and camera device changes between stereo images,corresponding points that are projections of the same scene point may have different RGB values of pixels and performance of stereo matching algorithms can be degraded.In order to improve the performance of algorithms,we present a stereo matching algorithm based on RGB proportional space.Firstly,histogram equalization is used for stereo images.Secondly,RGB proportional space is proposed based on the color formation model and pixel values are transformed into this space.Finally,disparity map can be obtained by the adapt support-weight algorithm according to color and gradient information of pixels in this space.Experimental results show that the proposed algorithm can obtain better disparitymaps under lighting geometry,illuminant color and camera device changes between stereo images.(2)In order to solve the problem that super-resolution restoration quality of the power line infrared image is affected by the camera shaking,a new super-resolution method combining SIFT feature matching is proposed.Take four adjacent frames for example.Firstly,an image is selected as the reference frame,and the transformation that map the reference frame to the other images is found by SIFT algorithm,which includes feature point detection and registration,removing mismatching pixels by RANSAC algorithm and image resampling.Then,take the reference frame and three resampling images into super-resolution reconstruction.The experimental results show that our algorithm can obtain a better super-resolution image compared with conventional super-resolution method.