Automatic Sorting and Mosaics of Unordered Overlapping Images Based on Fourier-Mellin Transforms and SIFT

摘要

Image mosaic is a key technology in the filed of image processing. An effective automatic sorting and mosaic method based on Fourier-Mellin transforms and SIFT is proposed and studied deeply. Firstly, the unordered overlapping images are sorted automatically by Fourier-Mellin transforms. Secondly, we use SIFT to extract feature points of two images and perform image registration based on feature vectors. Thirdly, image fusion is utilized to remove mosaic gaps. Some intuitive and persuasive simulation experiments are carried out by using the proposed automatic sorting and mosaic method. Experimental results demonstrate that the proposed method can automatically sort the unordered overlapping images and achieve seamless image mosaics. The proposed automatic sorting and mosaic method can be used in the fields of three-dimensional reconstruction, cooperative augmented reality and teleoperation robots.