Autonomous Image Georeferencing Based on Database Image Matching

As an experimental remote sensing microsatellite, LAPAN-A3/IPB satellite has been continuously producing daily multispectral images for monitoring Indonesian archipelago. The autonomous image preprocessing software has been developed to process raw image into systematically corrected image, except for high level image georeferencing process which is still conducted manually. This research aims to develop an autonomous image georeferencing for LAPAN-A3/IPB multispectral images so that entire image preprocessing can be executed autonomously. The algorithm is based on template image matching approach, by using numerous of previously processed images. The developed algorithm has been well validated and successfully georeferences LAPAN-A3/IPB multispectral images with a good accuracy. The time needed for the algorithm to process single image can be considered very fast, therefore the algorithm can be integrated into already established image preprocessing software. With this fully autonomous image preprocessing, each raw image captured could be processed inside one hour since its acquisition, allowing fast and consistent image distribution to the end user.