Improving Satellite-Aerial Image Matching Success Rate by Image Fusion

초록

Image matching is an important method to collect ground control points (GCPs) by finding correspondence between incoming images and chips of reference image maps. It is an essential process for automated precise geo-registration of satellite imagery. To get higher georeferencing accuracy, reference chips must be matched precisely on the images. The importance of higher matching success rate is increased with limited number of chips. In this study, we aim to match incoming satellite images against reference chips generated from aerial color ortho-images. Matching the two dataset is difficult since they have different spectral responses as well as different textures. We try to improve matching success rate by using pan-sharpened satellite images. The results showed higher matching success rate with pansharpened images due to similar spectral range and higher spatial resolution. Therefore, pansharpened image is helpful to improve image matching success rate in automated precise georeferencing of high-resolution satellite imagery.

키워드