Classification of Hyperspectral Remote Sensing for Production Minerals Mapping Using Geological Map and Geomatics Techniques

Abstract

The classification of hyperspectral images is an interesting job since the data dimension is huge for conventional classification procedures; normally several hundreds of spectral bands are attained for each image. These spectral bands can supported very rich spectral data of each pixel to find objects material .The objective of this research is to classify hyperspectral images for detection and production of detailed minerals mapping using geological map and Environment for Visualizing Images (ENVI) software. In this research, ASTER data and geological map have been used. Some techniques on these data are used such as enhancement, matching (linking), De-correlation, Band Ratio, stacking image and classification. The results showed that comparison of the two classification results showed the classification of stack image with the aspect and the slope provide more information than classification of ASTER image alone. Also, using ENVI software to generate 3D surface views.It concluded that capability of hyperspectral and its differentiation with multispectral data to extract detailed features from ASTER image.