Multi-sensor imagery processing

In considering applications of satellite sensors to exploitation of earth resources, it becomes apparent that imagery from a variety of sensors will be required to accommodate the various disciplines and the environmental constraints. To attempt crop surveys, for example, it appears that both the visual and the infrared spectrum will be utilized. This may be augmented by radar information to assist in defining water boundaries, regions, land with high water content, or to permit penetration over areas of continuous or seasonal cloud cover. One immediately becomes concerned by the lack of commonality in the data acquired by these sensors. The infrared imagery may be obtained by utilizing an optical line scanner with 20-30 channels covering the 0.3µ to 30µ band. The photographic imagery might be obtained by a panchromatic camera with 20-50 lines/mm resolution. The radar data could be provided by a side looking radar with several hundred foot resolution. The registration and resolution compatibility problems appear insurmountable, if one wishes to simultaneously utilize all the imagery to achieve the data classification. However, examining this in the light of information extraction as opposed to data processing one sees an immediate reconciliation. A specific sensor can be used as a primary sensor in establishing an initial categorization. Upon command the other sensory data can then be interrogated to provide supporting or verifying information.