The extraction of detailed and continuous information in remote sensing studies may have been by development hyperspectral images in the hundreds of band in recent years. These images have a high spectral resolution, but their spatial resolution is low. This constraintion creates mixed pixels from several materials in the image obtained from the site. In this study, we first used the fusion methods to increase the spatial resolution of the image, and then, in next step by using spectral unmixing methods to unmixed the spectral elements of the image. In this study, the spatial resolution of the input image, which combined images of ALI and ASTER from the Darehzar area, was increased by using a panchromatic image of IRS LiSS IV sensing with the spatial resolution 5 meters. In this step, we used appropriate and convenient methods such as PC sharpening, IHS, high pass filtering (HPF), Gram-Schmidt, Ehlers Fusion and wavelet transform combination methods based on principal component analysis and IHS color conversion to fusion the input images of area. In the next step, for the spectral unmixing, we used vertex component analysis (VCA) and N-FINDR algorithms to extraction endmembers of fused image. The resulting endmembers were used after the comparison with the USGS spectral library to identify the type of minerals associated with each endmember. Then, using linear spectral unmixing (LSU), using the whole mineralization associated with each endmember, the result is achieved in the area. For comparison and validation of the results, we used the 130 points were taken of the area for three alteration argillic, phyllic and propyllitic. In the discussion of the fusion of satellite images, the best results for DWT-PCA method were obtained, with a product accuracy of 84.61% and kappa coefficient of 0.7743. For comparison and validation of spectral unmixing algorithms of the endmembers, we used In this study, we used linear spectral unmixing method to unmixed based on the endmembers obtained from the extraction endmember algorithms. Based on the results obtained in the fusion, the best proposed method is DWT-PCA and in the spectral unmixing, the best algorithm to extract endmember is related to the N-FINDR method.