Image Registration And Fusion Research Based On Sift An Discrete Wavelet Transform Algorithm

With the development of science and technology, by its features of a large image perspective and high-resolution,image stitching has gradually went into the forefront of research in computer vision, image processing and computer graphics and so on. Image stitching technology, is what makes a series of images with the same scene overlapping parts together into sharp image with wide viewing angle, retains the basic information of the source image and no obvious resolution sutures. Two key steps of image stitching technology are image registration and image fusion. In this paper, using the classic matching algorithm-SIFT algorithm, and the wavelet transform image fusion as the main line, image stitching technology was conducted in-depth research.The main research work in the dissertation is as follows:(1) theory and basic technical processes of Image Stitching technologyThis article briefly described the basic process of image stitching technology, while introduced several common image transformation models. It focuses on the two key steps as image registration and image fusion technology. For image registration technique, this paper introduces the principle of image registration, image registration process and key elements, and common approaches of image registration.(2) SIFT-based image registration techniqueFor image registration technique, based on the classic SIFT algorithm, an improved SIFT algorithm was proposed. Including improved search space, and optimization of purification processes the registration point, which greatly improved the efficiency of registration and registration accuracy.(3) Wavelet-based image fusion technologyFor fusion algorithm of image stitching technology, this paper used the wavelet transform algorithm. By the analysis of the low-frequency high-frequency with fusion rules, an improved algorithm which produces smooth sutures in image stitching and achieves a seamless.