Study On Identification Mechanism And Information Extraction Of Rock And Mineral Based On Remote Sensing Spectrum

The extraction of rock-mineral information has become an important part of the application of remote sensing geology, with the deepening research on geological survey and the study of geological survey methods and techniques, the spectral characteristics of rock and mineral plays more and more important role in the identification of rock-mineral and the information extraction of them.The research relies on the project "the high-tech industry development project plans of Yunnan province (the application of mineral resources remote forecast evaluation technology in Yunnan Northeast mine set district), taking the rock-mineral samples collectted from Heqing region of Yunnan as research object, whose spectral curve can be obtained by the American ASD Field Spec 3 spectrometer and chemical elements content can be obtained by the Niton XRF analyzer.and then the correlative analysis on the percentage content of Fe element and the spectrum of rock-mineral is done by the technology of Spectral Derivative, concluding that the correlation between the percentage content of the Fe element and the bands of spectrum ranging from 1350 nm to 1449 nm is the largest, and then quantitatively extracted the characteristics of the spectrum parameters (absorption peak position, trough area, trough depth).At last, we established multiple regression inversion model of the percentage content of Fe element by taking the spectrum parameters as statistical variables using the statistical software SPSS. The inversion model shows that there exist correlation between the percentage content of Fe element and the spectrum parameters:trough area and the trough depth,which is extracted after the continuum is removed from the spectral curve, within the band range:from 1350nm to 1499nm.And we can predict the percentage content of Fe element by the trough area and the trough depth whose continuum is removed and the prediction error is less than 30%, which have a certain reference value.(1)In this paper the correlative analysis on the percentage content of Fe element and the spectrum of rock-mineral is done by the technology of Spectral Derivative, concluding that the correlation between the percentage content of the Fe element and the bands of spectrum ranging from 1350 nm to 1449 nm is the largest.(2) By using the mathematical methods, we extracted the spectrum parameters of the spectral curve and the spectrum parameters after the continuum is removed, then through correlative analysis, we conclude that spectrum characteristic factor associated with the percentage content of Fe element is:trough area, the position of the absorption band.(3)By the extraction of the characteristic parameters and the percentage content of Fe element as statistical variables, we use the statistical software SPSS, by stepwise regression analysis, established by the statistical variable of spectra information characteristics parameters.(4)Verify the percentage of Fe Inversion Model with the selected rock-mineral samples. The verification indicates that average error between the Fe element content of the rock samples calculated by the inversion model and the test value of the Fe element content used by the XRF is 19.7%, and the prediction results has a certain reference value.