Multispectral Papers

 
Research on Multispectral Imagery

With the availability of moderate resolution multispectral imagery, comparable in spatial resolution to aerial mapping imagery, opportunities exist to exploit the inherent spectral information of multispectral imagery to aid urban scene analysis for cartographic feature extraction and the construction of detailed databases for distributed simulation applications. Moderate resolution multispectral imagery with spatial resolution ranges of 5 to 8 meters can be collected with existing airborne multispectral scanners like Daedalus, ATLAS, AVIRIS, and MEIS.

Our research in multispectral scene information fusion utilizes moderate resolution airborne imagery and high resolution panchromatic aerial photography. Using traditional spectral classification techniques, surface material information is derived from the multispectral imagery, refined by monocular segmentations from the panchromatic imagery, and fused with high resolution stereo disparity maps.

This work has shown the feasibility of merging surface material information derived from moderate resolution multispectral imagery with estimates of height based upon stereo matching in high resolution panchromatic imagery. The goal is to use surface material information, normally highly correlated with object location in complex urban scenes, as a source of information for small scale mapping of man-made structures such as buildings and roads, as well as natural features such as soil, vegetation, and water. The fusion of height estimates with surface material estimates provides a unique synthetic three dimensional dataset that is not directly available in any airborne imaging sensor.

Additional ongoing projects in the analysis of multispectral imagery and multispectral scene interpretation include:

• Examination of probability, typicality, and discriminant images for verifying or modifying multispectral image pixel classification assignments,

• Utilization of existing broad area coverage cartographic databases with coarse spatial resolution for guiding the multispectral analysis and interpretation process toward local intensification of those databases.

• Fusion of collateral information from high resolution road network analysis and building detection systems with multispectral material classification.