Amir: a New Automated Multisensor Image Registration System

Current civilian/military surveillance and reconnaissance systems require improved capability to enable the coregistration of larger images combining enhanced temporal, spatial, and spectral resolutions. Traditional, manual exploitation techniques are not able to cope successfully with this avalanche of data to be processed and analyzed. Automated image exploitation tools can be used if images are automatically co-registered together, and thus ready to be analyzed by a further process. Such automated co-registration algorithms shall be able to deal with different civilian and military case scenarios, and helpful to be used successively in numerous applications such as image data fusion, change detection, and site monitoring. This paper describes the Automated Multisensor Image Registration (AMIR) system and embedded algorithms currently under development at DRDC-Valcartier. The AMIR system provides a framework for the automated multitemporal registration of electro-optic nadir images acquired from multiple sensors. The system is original because it is entirely automated and doesn’t rely on any user interaction to operate. Advanced image algorithms are used in order to supply the capabilities to register multitemporal electro-optic images acquired from different viewpoints, under singular operational conditions, multiple scenarios (e.g. vegetation, urban, airport, harbor, etc.), different spatial resolutions (e.g. IKONOS/QuickBird, Spaceborne/Airborne), and with a sub-pixel accuracy registration level.