Research on hyperspectral data mining technology of sandstone-type uranium deposit
Based on the theoretics of Hyperspectral data mining technology and the characteristics of imaging spectrum, this paper adopts field spectra and Hyperion satellite hyperspectral data to study the diagnosable spectral signature and quantitatively analyze and extract the absorption spectral feature of the sandstone uranium deposits in the north China. According to the spectral model, the spectra of uranium deposit can be divided into a series of Guassian simulated curves, and the absorption spectral signature can be quantitatively expressed by defined functions. Study shows that this approach is a useful tool for the spectral coding of database and spectral matching. Moreover, in terms of the information extraction approach and the processing method of extra-abundance data, the authors adopt high precision methods of radiance calibration and spectral mapping to study the geological feature of uranium deposits. Based on above researches, this paper discusses the prospect of Hyperspectral data mining technology in the exploration and evaluation of uranium deposits. Finally, the better results have been achieved in the practical application.
hyperspectral data mining geological mapping uranium deposit
Jielin Zhang Dechang Liu Daiyong Cao
Beijing Research Institute of Uranium Geology, Key Laboratory of Remote Sensing, National Defence Sc China University of Mining & Technology, Key Laboratory of Coal Resources, Ministry of Education, Be
国际会议
第八届国际矿床地质会议(Proceedings of the Eighth Biennial SGA Meeting)
北京
英文
110-113
2005-08-18（万方平台首次上网日期，不代表论文的发表时间）