Processing of large-area Quick Bird satellite imagery for the Kakamega Forest area (western Kenya)

In this article experiences are described gained during the processing of QuickBird satellite imagery covering an area of 631 km(2); the Kakamega Forest in western Kenya and its surrounding agricultural matrix. A correction of effects caused by atmosphere and orography (using ATCOR 3) has led to a considerable improvement of the original data. While atmospheric influences. e. g. a thin haze layer, could not always be minimized satisfactorily, orographic effects were successfully excluded. For mosaicing the two adjacent image swaths an elaborate procedure has been developed for the local geometric adjustment (making use of ERDAS Imagine and Leica LPS amongst others). In the resulting image the otherwise distinctly visible join line has become inconspicuous. Furthermore, ten pan-sharpening methods were evaluated upon their visual results as well as their preservation of the spectral characteristics. In a combined assessment the highpass filter method was evaluated as being best for the imagery at hand.