基于不同岩性背景的遥感影像蚀变矿物信息提取  

ASTER Image Alteration Minerals Information Extraction Based on Different Lithology Backgrounds

作　　者：于岩[1] 李建国[2] 陈圣波[1] 高学生[2] 路鹏[1] 黄爽[1] 张晨曦[1] 

机构地区：[1]吉林大学地球探测科学与技术学院,吉林长春130026 [2]天津地质调查中心,天津300170

出　　处：《地球科学（中国地质大学学报）》2015年第8期1391-1395,共5页Earth Science-Journal of China University of Geosciences

基　　金：国家高技术研究发展计划(863计划)项目(Nos.2012AA12A308;2008AA121100);中国地质调查局项目(No.1212011220469);国家自然科学基金项目(No.41402293)

摘　　要：为了减少由于不同岩性的反射率值差异而造成提取到错误的蚀变信息,采用先进星载热发射和反射辐射仪(advanced spaceborne thermal emission and reflection radiometer,ASTER)数据,以内蒙古昌特敖包地区为研究区,先开展岩性分区,再利用主成分分析和阈值分割的方法提取矿化蚀变信息,并进行了方法对比与野外验证工作.褐铁矿化蚀变信息验证点共12个,不分区直接提取的信息有6个与验证点吻合,分区后提取的信息有8个点与验证点吻合;绿泥石化蚀变信息验证点共5个,不分区直接提取的信息有2个与验证点吻合,分区后提取的信息有4个点与验证点吻合.研究结果表明,针对不同岩性存在的反射率差异,造成高背景值地区提取出较多的非矿致异常信息,而低背景值地区弱信息被噪声淹没的现象,该方法能够减少不同岩性产生的反射率差异.In order to reduce the incorrect alteration information extraction caused by the different reflectance values in the different lithologies, this paper explores alteration minerals information extraction by using ASTER （advanced spaceborne thermal emission and reflection radiometer） data, and taking Mongolia Changte＇aobao region of Inner Mongolia as the study area. The alteration information has been extracted by segmenting the different lithologies firstly, and conducting the principal component analysis and threshold segmentation methods. It is found that there are 12 limonitization information verification points, of which 6 points match the information extracted directly without segmenting and 8 points match after segmenting; In addi- tion, there are 5 Chloritization information verification points, of which 2 match directly and 4 after segmenting respectively. The results show that the reflectivity of different lithologies is different, resulting in extracting extra information in the high value background region and weak information submerged by noise in the area of low background values, and this method can effectively reduce the reflectance differences in different lithologies.

关 键 词：矿化蚀变信息 遥感 背景值 岩性分区 主成分分析 阈值分割. 

分 类 号：P627[天文地球—地质矿产勘探]