Automatic Ground Control Points Refinement For Remote Sensing Imagery Registration

Several sources of geometric distortion are contained in optical remote sensing images including earth rotation, platform movement, scanning nonlinearity. They result in geometric errors on scene level, image level and pixel level. It is critical to rectify the errors before a thematic map is generated, especially when the remote sensing data need to be integrated together with other GIS data. In this paper, image registration techniques are addressed and an automatic procedure is developed for refining manually selected ground control points. Both feature matching and intensity matching have been taken into account by generating three mapping functions to model the location relationships in two spatial directions and the brightness relationship between the master image and the image to be registered. The new method reduces the dependency on the accuracy of initial manually selected ground control points and improves the spatial correlation of the image. The registration quality is significantly improved. The developed procedure is demonstrated step-by-step using a Landsat ETM+ data set. The experiments showed that the proposed method is effective and easy to implement.