Researchers at the Fraunhofer Institute for Information and Data Processing have created the RecceMan (reconnaissance manual) interactive identification assistant to help the German army tell friend from foe on aerial or satellite photographs.
Cameras installed in aircraft and satellites can deliver a myriad of images of the Earth’s surface but they cannot provide an accurate evaluation of the pictures.
That assignment is left to trained analysts who have had to rely on an array of mostly printed reference materials for object identification as a search method that was optimised for image analysis was not previously available.
RecceMan reportedly provides analysts with hierarchically structured views of the signature characteristics of objects they wish to identify. The graphic representations and accompanying descriptions are maintained in a database, which is continually expanded and updated.
These ‘signature trees’ are tailored especially for the requirements of image recognition and enable the user to home in on identifying the object by referring to characteristics in greater detail and making use of predetermined questions.
‘Previously, the process was very time-consuming, since information had to be culled from all of the various reference materials. Now that data is consolidated, making it much quicker to access,’ said project leader JÃ¼rgen Geisler.
Under contract with EADS-Dornier GmbH, a pilot version of the program was integrated into the German Army’s CL-289 unmanned airborne reconnaissance system, where it is currently undergoing operational testing.
After this phase is completed, a decision will be made on how to use the software in other systems.