Extraction of SIFT Image Features for Efficient Image Registration

Abstract

This paper presents an image registration algorithm based on Scale Invariant   Feature   Transform (SIFT). SIFT is an image registration algorithm based on local features in an image. Compared to the previous registration algorithms, SIFT is more robust to variations caused by changes in size, illumination, rotation, and viewpoint of the images. The descriptors and key points that SIFT was able to gather demonstrate how resistant the algorithm is to scaling, noise, translation, and rotation. The image's most important details are first taken out. Later, dot products between the unit vectors are calculated to match the obtained points.  Finally, transformation matrix is obtained by applying RANSAC algorithm. Experimental results shows that the algorithm extracts the better key points, which can be used for used for image registration applications. Further work includes implementation of affine transform and image fusion techniques to obtain the registered image.