Rearch On Remote Sensing Image Mosaic Tchnology

This paper mainly studies the image mosaic technology. First, it introduces the image mosaic technology related concept, development status, the selected topic basis and UAV. Then the splicing layer theory and the splicing algorithm application and improvement are introduced, mainly focused on the image mosaic technology study.The mosaic object is remote sensing image, so a lot of research experiments are required to process remote sensing image, first , their region feature and feature point are extracted respectively, and then design the suitable method to determine desired feature extraction framework for splicing and calculate basic similarity, stitch image, at last match the image.This paper uses the Harris corner detection method for image matching feature extraction, the contourlet is improved so that it is effectively used in the Harris corner detection and overcomes Harris single scale defection to fuse image better.Before Harris feature extraction to image, a preprocessing is done to image. The region segmentation is applied to research on image mosaic, first, gray value of similar regional segmentation is processed, and Harris feature point extraction based on multi-scale is used. Remote sensing image splicing processing research is mainly reflected in that remoting sensing image uses improved Harris corner detection to extract feature, correct geometric, calculate the basic similarity. The simulation experiment comes last, experimental results show that this application and improvement of algorithm is feasible and effective.