A Study Of Mosaic Technique Based On Aerial Images

The aerial image plays an important role in environmental monitoring and geological disaster prevention.However,the coverage area of single aerial image is small.In order to get a wider range of ground information,image mosaic technique is needed.According to the research status of image mosaic technology,after the noise processing and detail enhancement in the use of median filtering algorithm and Sobel operator;Then,use the optimized SURF algorithm to extract the SURF feature points in the image;Subsequently,match the SURF feature points,and use the improved RANSAC algorithm to filter the matching feature points,as well as completes the solution of image coordinate transformation model;Afterwards,the image coordinate transformation model and the bilinear interpolation method are used to map the spliced image to the same coordinate system;The improved algorithm proposed in this thesis is used for image fusion at last.This algorithm improves the efficiency of image mosaic,the details of the processed image have no misplacement and blurred phenomenon,and there are no obvious stitching marks.What's more,the visual effect will be better and acceptable.