Gross feature detection in remotely sensed CAMS imagery using a multi-paradigm software approach

Abstract

Rules used to classify regions in images that have been remotely sensed from airborne or satellite platforms are not universal truths even for gross feature detection (i.e., trees, man-made objects, water, etc.). The classification of each pixel in a data set obtained from an aerial or satellite image can be a difficult and time-consuming process for both a computer and a human. Consequently, throughout the National Aeronautics and Space Administration (NASA) information systems researchers are seeking innovative approaches that will assist human experts in the time-consuming endeavor of classifying remotely sensed data. This article reports on a study, commissioned by the director of NASA's John C. Stennis Space Center (SSC) in Mississippi, to investigate the use of Artificial Intelligence (AI) techniques in the classification of remotely sensed data collected using the Calibrated Airborne Multispectral Scanner (CAMS). The study resulted in the development of a proof-of-concept software system that takes a multi-paradigm software approach to gross feature detection in remotely sensed CAMS data.