AN AUTOMATED GEOREFERENCING TOOL FOR WATERSHED SCALE FLUVIAL REMOTE SENSING

In recent years, remotely sensed image datasets with ground resolutions below 10 cm have seen widening applications in fluvial sciences with the development of methods for image based measurements of bathymetry, grain sizes and habitat types. However, given that these datasets typically contain hundreds or thousands of images, one of their key limitations is the need to georeference this large volume of imagery. Automated registration of remotely sensed imagery based on pattern matching algorithms has been the focus of much publication in other disciplines and there are now methods capable of automatically registering a newly sensed image to a pre-existing reference image. Based on such strategies, this paper presents an automated georeferencing tool specifically designed for fluvial remote sensing which uses a simple and well established pattern matching algorithm. This new method allows for very large image databases to be automatically georeferenced without the recourse to ground control points from the field and thus requiring minimal labour. Copyright (C) 2009 John Wiley & Sons, Ltd.