Evolutionary Algorithms for Sensor Fusion and ATR

The objective of this project is to develop a method(s) for detecting, segmenting, and identifying man made objects from background terrain from sensor fused data using Evolutionary Algorithms (EAs). Phase I will develop proof-of-concept implementation and demonstrate the capability to accomplish detection, segmentation , and discrimination or identification with the purpose of providing high probability of detection and accurate identification, with relatively low false detection and misidentification rates. Phase II will develop and demonstrate an EA based automatic sensor fusion algorithm for FLIR, TV and Ladar sensors to detect and identify tactical targets. The performance of the fused algorithm will be compared to performance of algorithms operating on a single sensor and will also evaluate the computational requirements of algorithm and developmental system and demonstrate the algorithm and prototype system in an operationally representative environment.