Imagery Automatic Extraction and Placement System

The objective of this project is to develop an imagery automatic extraction and placement system to retrieve and accurately position cultural features on geo-specific terrain. Feature extraction includes noise edge elimination by Gabor filter, edgeextraction, polygon simplification by the Hough Transform, and corner extraction. The innovative approach is to utilize Memetic algorithms to group the separate features from stereo pairs in order to recognize and measure cultural feature dimensions andplace cultural feature of correct dimensions and texture in terrain databases. Current high-speed digital signal processing (DSP) provides the opportunity for complex image processing and algorithm implementation. The Memetic algorithm presents thestatistical confidence of the results. The image-processing tool can be applied to medical diagnosis, automatic image-to map-registration, geodesy and environment monitoring.