Study On Panorama Stitching Technique Based On Feature Registration

Panorama is an image with a wide-angle view. It can be taken with special cameras; it can also be constructed through image alignment techniques from regular hand-held cameras. The technique to create panoramas is among the hottest topics in field of IBR (Image Based Rendering).The thesis focuses on study of panorama stitching technique based on feature registration. The steps of panoramic image stitching are emphasized. Virtual reality and its scene rendering methods, image acquisition, image cylinder projection, image registration and image blending are discussed. The main works are detailed as followed:(1) The research status of virtual reality is reviewed. The existing problems of panoramic image generation are analyzed. The flow chart of panoramic image generation is shown.(2) The meaning, characteristic, catalog of virtual reality are dissertated. Virtual reality scene rendering methods are analyzed and compared. The thesis uses image-based rendering method that is explained in detail. Image-based rendering method and the characteristic of image-based rendering panorama video are introduced.(3) Image acquisition methods in the process of panoramic image generation are researched. The problem and requirement of image acquisition are explained. The aim of image projection and panoramic image types are analyzed. Cylinder projection model becomes the first-selected one in the thesis, which is explained. According to existing cylinder projection algorithm, model construction and coordinate transfer are derived and proved. The change law of image size after projection is discussed and the algorithm is tested.(4) The current image registration methods are introduced and their advantages and disadvantages are analyzed. The thesis uses the image registration method based on the future of the image, which is of better result and less calculation. The SIFT algorithm is improved, and the RANSAC algorithm is used. Compared with the original algorithm, the new algorithm increases real-time performance. The experiment improves the result of the algorithm, and the stitching images are shown.(5) In order to remove seam in the process of stitching, the study is done on blending methods. A linear interpolation is used to solve the problem.