Research On Image Mosaic Method And Its Application

The image mosaic technology has been a hot research direction of computer vision, image processing and computer graphics. It can be used to create high-resolution images of the large viewing angle, and has a wide range of applications in the field of virtual reality, medical image processing, remote sensing technology and military fields.This artcle analyses the background and meaning of image mosaic, researching status,related algorithm. Combining with the process of image stitching technology, it researches the way of image acquisition and the model of image transformation. Analyses and compares image registration and the relative merits, focusing on the ways of image registration which bases on features. Introduced the principle and relative merits of Harris, SIFT,ASIFT feature extraction operator.From the experiment result, ASIFT is of full affine invariant. In the process of initial feature points matching, compared exhaust algorithm,k-d and BBF algorithm.In the exact matching process, compares robust estimation algorithm such as M-estimators,RANSAC estimators,LMedS estimation method,and proposes using Guass-Newton algorithm to optimize LMedS,so that we can purify feature points. On image fusion, compares the average superposition method, weighted average method and multi-band and blending.Through simulation experiment, we can learn that the multi-band and blending can realize seamless stitching of images.At last introduces it’s application on digital campus and other fields.The experimental results show that the improved algorithm solves the limitation of the image feature extraction when we get the images the by hand-held camera.It can mismatch feature points by noise, light effects and improve the capabilities of feature extraction and matching precision, the resolution of synthesized image is high.