Imagery Automatic Extraction and Placement System

This Phase II project is aimed at developing an automatic, imagery based, processing software that allows extraction from and placement of culture features (such as, buildings) in images of a database. Object classification is effected through features based on edges and intensity distribution. Additional considerations involve noise suppression and edge aggregation. The innovative approach is to utilize imagery derived from imaging sensors, including stereo cameras, in order to effect classification of culture objects through a set of features and isolate them from the surrounding background. Of equivalent emphasis is the derivation of a geometrically and radiometrically consistent approach for the placement of culture objects of the correct extent and texture in existing databases.