Advances in Biomedical Engineering, cilt.14, ss.20-25, 2012 (Hakemli Dergi)
The main objective of this study is to test performance of pan-sharpening algorithms on image classification accuracy. The study is consisting of two main steps. In the first step pixel level simple pan-sharpening and IHS pan-sharpening algorithms are applied to Quick-Bird and Ikonos images to integrate the geometric detail of a high-resolution panchromatic (Pan) image and the color information of a low-resolution multispectral (MS) image to produce a high-resolution MS image. In the second step pan-sharpened images are classified using MCL classification procedure and performance of these algorithms on the urban environment classification is evaluated by using error matrix of classified data. The result of the study indicates that performance of pixel level pan-sharpening algorithm for urban class detection is considerably higher than IHS based pan- sharpening algorithms in both Quick-Bird and Ikonos images.