Information Extraction Of Rich Potassium Rock Based On Aster Remote Sensing Data

Water-insoluble potash resources are very abundant in our country, various potassium rich rock mainly of potassium feldspar and illite has a very wide range of distribution, almost all over the country. If this part of potential resources can be used effectively, to a certain extent, the lack of water soluble potassium resources can be made up for accordingly. Shortage of potash resources in China, and to promote sustainable development and utilization of potassium(fat) mineral resources, advanced remote sensing technology is adopted to carry out the remote sensing geological survey and evaluation of water-soluble potassium mineral resources in Lushi and Song county, henan province. Research purpose is to obtain the objective based data of working area and form a comprehensive analysis and resource potential evaluation report, to provide certain technical support for obtaining remote sensing exploration data of potassium rich intrusive rock belt in the north rim of east Qinling quickly and accurately.Remote sensing in the application of rock information extraction mainly relies on band synthetic, image enhancement and spectral features of rocks. In recent years, the quantitative inversion of oxide content in the rock for the analysis of rock types and identification of the lithology is a direction of quantitative inversion in remote sensing petrology. The chemical analysis result of rocks shows that through the oxide content to identify rich potassium in rock is an effective method. In order to identify rich potassium resource rock more accurately, rich potassium rock mass in Song county, Henan province is selected as the research object. Spectrums of 37 rock samples were determined by ASD portable spectrometer, based on rock identification theory of remote sensing petrology, by using the statistical analysis based on image, and rock spectral features, lithologic identification of the potassium rich rocks in the study area was conducted, and the analysis and comparison of identification results of the two methods were done. Then, based on spectrum transformation data and rock oxide content of the sample point in the study area, oxide content inversion model was gotten by multiple stepwise regression analysis method. Using ENVI image processing software and ASTER remote sensing data, the inversion of the oxide content in the study area was processed. Through IDL, CIPW(standard mineral calculation) algorithm was gotten. So mass fraction of main standard minerals, quartz, albite, orthoclase and calcium feldspar of rocks in the study area was gotten. Finally on the basis of arithmetic operations and the ratio of operations, the extraction rules of the rich potassium rock was built. Rich potassium rock mass in the study area was extracted.Comprehensive analysis of quantitative and qualitative recognition results, comprehensive analysis and evaluation of extraction method of the rich potassium rock mass in the study area was processed. The experimental results show that the method that based on the measured spectrum, using rock spectral curves and chemical analysis results to set up rich potassium rock information extraction model, has fast recognition potential of rich potassium rock resources, which should be widely used.