基于Hyperion高光谱数据的矿物蚀变提取——以内蒙古西部狼山地区炭窑口矿床为例

以已知的矿床——内蒙古炭窑口矿床为研究区,基于围岩矿化蚀变理论,利用高光谱遥感技术对矿化蚀变异常信息进行提取。首先对Hyperion影像进行预处理,包括去坏线、条带、大气校正及光谱重建。其次,从反演后的反射率数据提取5类端元矿物,通过分析矿物组合,最终确定5类矿物并利用波谱角制图法在研究区填图,进而提取矿化蚀变信息。在以上研究过程中改进了去除条带的方法。在获得蚀变信息后,结合炭窑口研究区的地质图进行应用分析,通过解释异常,验证其是否为矿化引起的蚀变。最后,再根据研究区的地质情况和前人对矿床的研究成果,尤其是对含矿地层的研究成果,结合提取的蚀变矿物,预测其他有利靶区。结果表明,获得的矿化蚀变信息与矿点吻合较好,具有一定的借鉴意义。

Mineralizing alteration extraction based on Hyperion hyper-spectral data-taking Tanyaokou deposit, Inner Mongolia, China as an example

Taking known deposit called Tanyaokou as an experimental region and based on the theory of wall rock alteration, study and experiment have been carried out on mineralization alteration extraction with hyper-spectral remote sensing image. Specifically, research is started with finding some indicative symbols based on alteration theory and studying spectrum features. Furthermore, pre-treatment of EO-1 Hyperion imagery includes remove band lines, de-stripping, atmospheric correction and spectral reconstruction. Five end members have been extracted from inversion reflectance, which are used to mapping the study area using SAM method. Therefore, five alteration anomalies have been extracted. Besides, overlay analysis of hydroxyl anomaly information extracted from multi-spectral remote sensing imagery with alteration minerals distilled from hyper-spectral data in order to verify the correctness and feasibility of extraction, at last, repeatable technology processes of extraction for mineralization alteration using multi-source remotely sensed imagery have been established. Finally, according to geological setting and previous research about ore deposit especially on ore bearing stratum, the authors forecast other favorable target areas using alteration distribution map. The result reveals that extraction of anomalies agree with the ore-bearing strata, which shows practical meanings.