Truth Quest: Enabling Operational/Exercise Data

The Phase II project will address enhancing ATR by developing techniques that use single and fused remotely sensed data sets to precisely characterize terrain environments, areas of potential trafficability, and areas of EOC clutter that can be ignored bythe ATR algorithms. These precision EOC identification and clutter-removal processes can be utilized as (1) a pre-ATR (terrain/ target potential or possible area) assessment and/or (2) a post-ATR refinement (change detection) process. The toolbox willconsist of multiple processing and analysis pathways suited to a variety a data sets and conditions. Data analysis will rely upon the use of spectral and spatial analysis. The analysis includes single image classification, fused classification, andsingle data type and fused data spectral/backscatter and spatial analysis techniques The process precisely discriminates and classifies fine scale landscape features. The use of spatially explicit techniques for classification and spatial patternidentification assists in (1) assess classification uncertainty; (2) quantification of expected rates for false negatives and false positives; (3) provides a probabilistic assessment of landscape features for both target and non-target classes. Theresult is a process to measure landscape features (characterize terrain) with a high level of precision and accuracy.