Advanced Multiscale & Multiresolution Image Enhancement and

The proposed STTR project (Phases I and II) will develop multiscale and multiresolution algorithms and software, as well as produce a simulation testbed for evaluating advanced techniques for image enhancement and classification using sensor data from multiple UAVs and UCAVs. For development and testing we will utilize multi-sensor data such as SAR, ISAR, IR, MMW, and Ladar, obtained from both simulation packages (eg. IRMA, XPATCH) and from real data (eg. the MSTAR dataset, ARL dataset, ONR Passive MWM dataset). During the course of the project, synergies between enhancement and classification methods will be identified, and the most promising techniques will be selected for in-depth testing and commercial development. Prof. Coifman from Yale Computational Mathematics group will provide an Algorithm Development Toolkit, including an extensive collection of state-of-the-art wavelet-based tools for enhancement and classification. Phase 1 effort will consist of the following specific tasks: 1. Selection of imaging data for algorithm development and testing; 2. Development of a toolkit consisting of multiresolution and multiscale algorithms for image enhancement and classification; 3. ATR performance evaluation and selection of ``winners'' (optimal pairs of image enhancers/classifiers); 4. Testing of the selected algorithm(s) using real and simulated datasets; 5. Development of an ATR component for the UCAV Simulation Testbed; 6. Final Report. 7. (Option) Functional integration of the ATR component into the UCAV Simulation Testbed.