Research On Construction Of Rock Spectral Database And Intelligent Classification In Xingcheng Area,Liaoning Province

With the exponential growth of remote sensing data,how to manage and analyze remote sensing data has become the most important role of remote sensing technology in scientific research,economic production,disaster prevention and control.The establishment of a hyperspectral database is conducive to the dataization of hyperspectral information.It is convenient to manage the hyperspectral information data.And the hyperspectral information database can make data analysis easier.It also makes it possible to apply big data ideas to hyperspectral data.Using the Python programming language,the Rock Spectrum Assistant(rock mine hyperspectral data management and drawing assistant software)was written to classify and manage the collected typical rock spectrum data of Xingcheng Area,Liaoning Province.This database realizes the function of drawing hyperspectral data of rock and mine.And it can carry out basic addition,deletion,modification and checking functions.Compare the number of spectra of multiple rocks plotted.Analyze the characteristics of the three rock spectra.On this basis,three types of sandstone,limestone and granite are classified by comparing SVM,KNN,RF and DT.In general,the support vector model performs best when it comes to classification.The classification accuracy reaches 98.8%,and the fastest time is only 1.02 s.So this database selects the support vector machine model.Then use the Rock Spectrum Assistant to store the collected rock spectrum inXingcheng Area and classify it at the same time.