论文部分内容阅读
由于传统的岩性分类方法受岩石辐射干扰因素大,存在“同物异谱”以及“同谱异物”现象,岩性分类精度低,所以在深入分析岩石矿物光谱特征基础上,以西昆仑成矿带地区的二长花岗岩、石英正长岩以及正长岩为研究对象,基于这3种岩性的实测光谱数据以及先进星载热发射和反射辐射仪(advanced spaceborne theemal emission and reflection radiometer,ASTER)影像数据的波段设置特征,建立了RI和SI两种光谱指数.利用所建立的RI以及SI光谱指数对ASTER遥感数据进行岩性分类.结果显示,RI和SI两种光谱指数法在提取二长花岗岩时精度达到70%以上,石英正长岩精度为80%左右,与最大似然法得到的分类结果相比,这两种岩性的分类精度明显提高了.
Due to the large interference factors of rock radiation, the traditional lithology classification method has the phenomena of “same matter different spectrum” and “same spectrum foreign object”, and the lithology classification accuracy is low. Therefore, on the basis of deeply analyzing the spectral characteristics of rock minerals, Based on the measured data of these three kinds of lithology and the advanced spaceborne theemal emission and reflection radiometers (Reflectance Spectroscopy), the authors studied the monzonitic granite, quartz syenite and syenite in west Kunlun metallogenic belt. (RI) and SI (subscript SI) spectra were established for the lithology classification of ASTER remote sensing data by using the established RI and SI spectral indices.The results showed that the RI and SI spectral indices The accuracy of the extraction of monzogranite reaches more than 70% and the accuracy of quartz syenite is about 80%. Compared with the classification results obtained by the maximum likelihood method, the classification precision of these two lithologies is significantly improved.