A Fuzzy Logic Method for Extraction of Geographic Objects from IKONOS Imagery

A high resolution satellite image is characterized by detailed geographical information offered. However, the heterogeneity disrupt a processing image, whose the principle object is transformation into geographic card; which is until now carry out by classic operations. The present  paper propose a new methodology of geographic objects identification from multispectral image, which combine between colored composition and radiometric characteristic of geographic classes, the mathematical morphology in order to solve over segmentation; the most problem of high resolution imagery, followed by an oriented segmentation and a fuzzy logic classification based on the formulation of some observations into decisions rules. We focus our study on IKONOS high resolution images on some areas of Algeria.