Hamid Reza Bakhtyari - Georg-August University, Inst. Of Forest Assessment and Remote Sensing Geottingen, Germany
Ali Darvishi Boloorani - Georg-August University, Dep. Of Cartography, GIS&Remote Sensing, Geoettingen, Germany
Mozhgan Abasi - Tehtan University, Natural Resources Faculty Karaj, Iran

Hyperspectral imaging can be defined as acquisition of images in hundreds of registered, contiguous spectral bands such that for each picture element of an image it is possible to derive a complete reflectance spectrum (Goetz, 1992) . Accurate estimation of biophysical and biochemical characteristics of vegetation for agricultural and forestry purposes is one of the main inherent abilities of hypersepectral images. The main focus of this study is to evaluate EO-1.Hyperion abilities foe plan conver mapping over the city of Bam in southeast of Iran. In these respect two main classification techniques included Spectral Angle Mapper (SAM) and Object-Oriented Classification (OOC) have been used. The results show that these classification methods demonstract hight ability of using EO-1-Hyperion satellite data for mapping Plam trees. Both the SAM and Object Based approaches have a good potential to use the rich spectral properties of Hyperion imagery. The object based method was especially effective in identifying smaller objects and accuracy assessment results show the Object Based classification give an Overall Accuracy and Kappa coefficient . 87.93% , 96.6% , respectively

EO-0 Hyperion , Hyperspectral , Object Based Classification , Plan, Remote Sensing , SAM