Hyperspectral Mineral Identification Based On Modified Gaussian Model

Hyperspectral remote sensing is one of import high-tech and technologies in the remote sensing area, especially in Geo-remote sensing area form home and broad.Due to the limitation of the spectral resolution in traditional Multi-spectral remote sensing, it is complicated to recognize the special but rich spectral information of ground objects. From the 80s'of last century, hyperspectral remote sensing is wide used in mine search, the extraction of the spectral information and so on, also we get so many successful results. The research shows that, it is possible to extract the indicative spectral information of the geological body with the assistant of Nano-hyper remote sensing data.Automatic identification of rocks and minerals on hyperspectral remote sensing images is an important aspect of geological applications of hyperspectral remote sensing. Many arithmetical methods have been developed, one of which Modified Gaussian Model is developed and applied by many researchers this years. The goal of this paper's research work is to implement the function of automatic identification of rocks and minerals on hyperspectral images using Modified Gaussian Model. The research work was divided into three parts. Firstly, Implement the function of input, extraction, and ouput of the spectral liboratory data and bands selection. Secondly, Fit the spectral curves of minerals with the Modified Gaussian Model. Lastly, Identify the rocks and minerals on hyperspectral images and output the identification results.After analyzing and comprehending the spectral character of mineral and rock deeply, this paper has mainly done many fitting analysis on the diagnostic spectral character of alunite, chlorite, goethite, hematite, illite and kaolinite using Modified Gaussian Model with the IGCP spectral library data. And with the model results some identification work of the six minerals has tried to do on the hyperspectral remote sensing data, some good results has got. The whole research work indicated that it is an useful and feasible method for hyperspectral remote sensing application on geology to fit diagnostic spectral character of minerals and rocks and to do minerals and rocks identifications on hyperspectral remote sensing data.