The development of hyperspectral sensors is the most significant recent breakthrough in remote sensing. Hyperspectral remote sensing is widely used in geology in that it can provide ample spectral information to identify and distinguish spectrally unique mineral. Hyperspectral imagery provides the potential for more accurate and detailed information extraction than possible with any other type of remotely sensed data. Spectral angle mapper (SAM) algorithm has been successfully used in geological mapping in many years. In conventional SAM method the remote sensing image should be atmospherically corrected at first. It assumes that the remote sensing data have been reduced to apparent reflectance, with all dark current and path radiance biases removed, and then calculates the spectral similarity of image spectra to reference spectra which can be either laboratory or field spectra or extracted from the image. This paper introduces a simplified method using SAM algorithm. The remote sensing data is converted to radiance and transformed into surface and atmospheric reflectance directly without atmospheric correction. The known mineral reflectance spectrum is also transformed to surface and atmospheric reflectance by MODTRAN (MODerate resolution atmospheric TRANsmission) model, then SAM algorithm computes spectral angles between the surface and atmospheric reflectance of the reference spectra and the image spectra without atmospheric correction. This paper uses the hyperspectral remote sensing image located at Beiya gold deposit in Yunnan province in southwest China, November 11, 2004. The main mineral reflectance spectra were acquired by the handhold spectroradiometer in situ. In order to validate die result, the SAM algorithm using the field spectrum matched with the atmospheric correction hyperspectral data was used. The correlation coefficient between the result rule image of using the simplified and conventional SAM method is 0.78. To compare the mapping result with the geological map and previous research, they are consistent with each other. It proves the method is effective. Comparing these two approaches, the former matches two kinds of surface spectra data without atmospheric influence, and the latter matches two kinds of surface and atmospheric spectra data contaminated by the same atmospheric condition, nevertheless, the latter is simpler. It is so complicated process of atmospheric correction of remote sensing image that it will consume a lot of time to remove the influence of atmosphere successfully. However, in the latter, only several mineral reflectance spectra attenuated by the atmosphere when they transmit to the satellite can be calculated, so it improves speed and precision of operation greatly.
展开▼
