|
This paper discusses RADC/USAF Research and Development activities in the areas of image processing and target identification relating identification and location of ground targets in a tactical scenario. The primary objective is to reduce the reconnaissance cycle from days/hours to minutes and seconds commensurate with the near real time (NRT) intelligence requirements of the tactical forces. This NRT intelligence is essential to support strike of increasingly mobile enemy weapon systems. In currently fielded systems exploitation of film based reconnaissance is extremely slow, greatly lagging collection rates. There are three essential elements to a NRT tactical intelligence system. They are NRT imagery collection, NRT air to ground image data link and NRT imagery exploitation. The thrust of this paper will be to discuss technology required to support the later, NRT imagery exploitation. Technology intensive efforts are categorized under each of the exploitation elements (target detection, identification and precision location). R&D in the area of target detection consists of exploratory and advanced development work units in automated target correlation, automatic change detection and pipeline image processing for screening probable target areas. Target identification R&D to be presented includes automatic techniques for pattern recognition as well as semi-automated techniques for aiding an analyst by correlating various sensor and intelligence inputs to permit target identification. Near real time precision target location techniques will include techniques for locating imagery targets in a predefined precision photographic data base as well as techniques for performing location simultaneously with target identification. Summary and conclusions for this presentation discuss commonality aspects of a digital image exploitation system relating to a "universal" image exploitation system and the potential for its use in NATO. Potential areas for developing cooperative R&D programs in support of this universal system objective will be identified.
|
