【作者】 张守林；

【导师】 王润生；

【作者基本信息】 中国地质大学（北京） ， 地图制图学与地理信息工程， 2006， 博士

【摘要】 多光谱遥感矿化蚀变信息提取,一直是遥感地质研究的热点和难点,虽然在定性与定量方面总结出了许多技术方法,但对于缺少基础地质资料的工作区,却没有完善的方法。如何直接利用多光谱遥感数据来圈定蚀变区段,从而指导地质找矿勘查工作,这的是一个亟待解决的问题;遥感植被指数应用在农业和林业调查中非常有效,能否研建出一种矿化蚀变因子,直接圈定矿化蚀变地段,该问题的解决必将给多光谱遥感在地质找矿中的应用带来突破性进展;与此同时,传统的遥感影像及蚀变信息表达往往不符合人们的视觉习惯,能否研建出一种三维显示效果逼真、符合视觉习惯的正地貌遥感影像表达方式,此问题的研究也将会为遥感技术增添新的活力与生机。以上技术难点是本论文研究的主要内容。本论文利用ETM遥感数据,采取从已知到未知、从理论分析到方法实验的技术思路,选择地表覆盖较少的新疆东天山和阿勒泰地区以及智利的银山进行矿化蚀变信息定量提取研究,取得了以下主要成果:(1)通过对蚀变矿物、岩石波谱特征和蚀变信息提取方法的机理研究,明确提出了铁化因子、泥化因子、综合蚀变因子的概念:IFE=ETM3/1;IOH=ETM5/7;I综合=ETM(5+3)/(7+1)(2)依据遥感数据统计分析,总结出根据蚀变因子的均值、偏度、标准偏差确定蚀变异常下限值的定量计算公式,解决了在地质资料空白区遥感蚀变信息的定量提取这一技术难题;(3)由于遥感蚀变信息为不同蚀变矿物(岩石)的综合反映,因此在蚀变分级上既要考虑蚀变因子数值的大小,又要考虑不同蚀变矿物的集合特征。本文通过对蚀变(异常)经对分级、分类方法的对比研究,认为动态聚类法和最优分割法是比较理想的两种分级方法;(4)因“沙漠漆”、岩石表面“氧化铁膜”、正常沉积的粘土、植被等非矿化异常的存在,故需对遥感蚀变异常进行筛选。本论文总结出了基于遥感综合解译法、多源信息分析法、野外验证法三种异常筛选方法;(5)在分析遥感成像机理和视觉习惯的基础上,本论文研究实现了遥感影像及蚀变信息符合视觉习惯的正地貌表达,且该技术方法已经获得了发明专利。更多还原

【Abstract】 The extraction of mineralization information based on mutilspectral data has been hotand difficult for years. Although there are a lot of techniques summarized both qualitativelyand quantitatively, there is no complete method aiming at the target areas short of basicgeologic data. How to delineate an alterated zone directly through mutilspectral data to directgeologic prospecting is an important issue to be solved. The application of vegetation indexesto agriculture and forestry is proved to be very effective;therefore there may exist the samekind of mineralized alteration factors which can be used to delineate an alterated zone directly.If this issue can be solved, it is certain to result in a breakthrough in mutilspectral data’sapplication to mineral prospecting. Meanwhile, human’s visual habits require a more vivid3D positive landform images rather than the traditional remote sensing images and alterationinformation, which is also a very important issue. The research of this paper centers aroundthe above issues.In this paper ETM data of the sparsely covered east Tianshan Mountains and Altai areain Xinjiang as well as Silver Mountains in Chile were chosen to extract mineralized alterationinformation based on a method from the known to the unknown and from theoretical analysisto practical experiment;as a result the following was achieved.(1) Concepts of ferrous mineralization factor, argillation factor, and comprehensivealteration factor were invented through the research on the mechanism of alterated minerals,rock spectral characteristics, and alteration information extraction.IFE=ETM3/1;IOH=ETM5/7;I 综合=ETM(5+3)/(7+1)(2) According to RS data statistical analysis, a formula to calculate the lower rangevalue of alteration anomaly based on mean, bias, and standard deviation were summarized,which solved the issue of quantitatively extracting alteration information by using RS datain an area without existing geologic data.(3) RS alteration information is an integration of different alterated minerals (rocks);therefore both the magnitude of alterated factors and the collective features of differentalterated minerals should be taken into consideration for alteration classification. In thispaper through the study on classification and categorization of alteration (anomaly), it’sconsidered that the dynamic clustering method and the optimal partitioning method are themost ideal ones.(4) Due to the existence of such non-mineralized anomalies as “desert varnish”, “ferricoxide films” on rock surface, normally deposited clay, and vegetation, screening is requiredfor RS alteration anomalies. In this paper three screening methods are put forward, whichare the comprehensive interpretation method, the multiple-information analysis method,and the filed verification method.(5) Based on thorough analysis of RS imaging mechanism and human’s visual habits,in this paper a positive landform representation method was invented, which has beenpatented.更多还原