The combinatorial explosion: defining procedures to reduce data redundancy and to validate the results of processed hyperspectral images

This paper will focus on ways to manage large numbers of remotely sensed data images as results of image processing, a primary problem especially for those dealing with hyperspectral images that pose considerable issues due to the elevated number of channels. While briefly introducing the results of the application of several common image processing techniques in the target area of Aquileia (NE Italy), the current paper will discuss the necessity to define a set of procedures to reduce the number of final images to be used for visual inspection, selecting the ones that do not carry redundant information. Consequently, cross process coverage, detected traces evaluation and process validity tables will be presented and the results of their application discussed in order to provide information about how to reduce the images to a small number and be able to insure the complete coverage as regards to the detectable traces.