Conclusions

Pattern and texture recognition between vegetation types or conditions will likely continue to be difficult. It seems that many of the GIS analytical techniques were designed for work with bands of spectral information created by satellite imaging devices. Kite aerial photographs are a close up view of what satellites may try to capture from space. The imaging devices currently available for KAP produce either digital, color-film or color-infrared imagery. Fortunately modern photo editing programs enable us to manipulate the imagery from these media for GIS Analysis.

It required several hours of trial and error to find techniques that would convert these digital RGB images into something that could be analyzed by Idrisi32. The approach that I chose for this project worked well for a simple delineation pattern. It could be cumbersome if there were more variables to delineate by hand. However the filtering and channel separation techniques in a photo editing program would suffice in many cases without hand delineation of boundaries.

I am confident that the time that I invested in this class project will be useful in achieving the purposes of the EPSCoR project. During this process I also gained insights as to how these techniques could be developed into powerful research tools to study vegetation patterns and other sorts of similar-colored scenes.

next page

home