Study On Panorama Unrolling Based On Fisheye Lens

Image shot by fish-eye lens has more information than normal cameras. And the angle of the view field of the fish-eye lens is bigger than normal cameras. For this reason, using fisheye lens can get scenery in a large field, simplifies the progress of image acquisition, and save the hardware resources. Fish-eye lens has many applications in many aspects. However, the pictures taken by it have great distortion, it has to calibrate before observing. After calibration, the image is more suitable for human vision and can get more information.Based on analysis of fish-eye image-forming principle and the models, this paper proposes a new fish-eye correction scheme. The main plan of this paper is to use different calibration methods to correct photographs taken by fisheye lens, in order to achieve the best results suit for human observing.In the experiment, this paper uses the unique imaging features of fisheye lens that leaflet picture contains a large amount of image information. In the research, we correct the image from the fisheye lens, then we finished the image registration and image the function of the system was realized and each module ran successfully. Using the Matlab to check the algorithm and the progress, and generate the panorama successfully.The ending of this paper is the result of the experiment. It confirms the plan proposed in the paper is practical and effective.