An efficient method to match large images

Image matching is one of the active areas of research in today's world. There exist many ways in which two images can be matched. When the size of the image is very large, it incurs more computational time and cost. In this paper, an efficient algorithm to match large images is proposed. This algorithm works by generating key features using SIFT algorithm. This SIFT features are then clustered using k-means clustering algorithm The output of k-means is then subjected to coupled decomposition phase. This will result in a set sub-images of the original images. Subsequently matching is done by finding the line segments in the corresponding sub-images. If all the sub-images are matching then the two images are considered to be fully matching. In the last part of the paper, the full image matching method is modified so that it can be used to detect partial match as well.