Downloads & Free Reading Options - Results

The design and development of multispectral remote sensor systems and associated information extraction techniques should be optimized under the physical and economic constraints encountered and yet be effective over a wide range of scene and environmental conditions. Direct measurement of the full range of conditions to be encountered can be difficult, time consuming, and costly. Simulation of multispectral data by modeling scene, atmosphere, sensor, and data classifier characteristics is set forth as a viable alternative, particularly when coupled with limited sets of empirical measurements. A multispectral system modeling capability is described. Use of the model is illustrated for several applications - interpretation of remotely sensed data from agricultural and forest scenes, evaluating atmospheric effects in LANDSAT data, examining system design and operational configuration, and development of information extraction techniques.