Study Of Image Registration Algorithms And Implementation Technique

Image registration,which is a fundamental problem in image procession, Imageregistration technique has been used in computer vision,medical inspection andmajorization, image mix together fields.Scale Invariant Feature Transform（SIFT） is very broadly used in the seale androtation in the image changes，it also has a strong adaptability in image deformation.In thesame time, this feature is easy to follow-up the match picture and has high identify abilityto match. traditional image registration which is throughed the use of these characteristics,has got a large degree of improvement in many of the common problems, SIFT is a kind ofautomatic registration important ways of Image feature extraetion and image matching.This paper first introduces the image evolution and a variety of algorithms,thenfocuses on a high-level characteristics way in image registration (SIFT algorithm). Itincludes the history of SIFT algorithm, the algorithm superiority, and primary process,amethod of image matching is proposed by combining wavelet transform with Scaleinvariant Feature Transform.To increase illumination and discriminative power, the SIFTfeatures of adding color information have been well developed. The low frequency imagecan be obtained through wavelet decomposition from both the original image and themodel image. And then features with color SIFT are detected using SIFT algorithm. Thematches are constructed by the optimized implementation of BBF algorithm, and verifiedby RANSAC. The algorithm can reduce computation cost of image matching and improvethe speed, besides, it is invariant to image change.