Automated Feature Extraction Capabilities for the Development of High-Resolution GEOINT Feature Data and Constructing Correlated Databases

Data resources available for automatic feature extraction (AFE) have expanded significantly in the last few years. Available sensor data now includes high-resolution multispectral and hyperspectral sensors, synthetic aperture radar (SAR), and accurate height measurement sensors such as LIDAR and interferometric SAR (IFSAR). Current AFE tools are unable to fuse and process all the new types of sensor data. TSC proposes a program for AFE algorithm development and testing that ingests all available sensor data, and extracts buildings, structures, road networks, and waterways. The method uses a knowledge-based expert system (KBES) that is trained with initial user interaction. The KBES methodology facilitates the fusion of multiple sensors, and also adapts easily to different combinations of sensors. TSC will develop algorithms for extracting vector features from the input raster data, and creating an output in a vector map format