【作者】 吴德文；

【导师】 刘少峰；

【作者基本信息】 中国地质大学（北京） ， 地学信息工程， 2006， 博士

【摘要】 当前地质矿产勘查工作已从传统找矿进入到高科技应用时代,遥感与地质、地球物理、地球化学勘查手段相结合的综合找矿方法成为现代找矿的主流。面对我国矿产资源供需矛盾日益突出的现状,地学数据资源的利用率仍然较低以及数据处理水平相对滞后的状况,开展本项研究工作无疑对提高矿产勘查技术水平、提高找矿效果和效率具有重要的现实和理论意义。本文以成矿区带的矿产资源潜力评价为目标,利用区域地质矿产资料、多光谱遥感数据(TM、ETM、ASTER)、区域化探数据、区域航磁和区域重力等多源数据,进行信息处理、建模、提取和综合挖掘等技术方法研究,为矿产资源定量化综合评价提供有效的技术手段和新的信息源。研究内容包括:遥感与地面观测数据的转换、整合与集成技术,遥感与地面观测数据的找矿信息提取技术,多元地学数据的综合处理技术,以及多元地学信息的矿产资源潜力评价应用技术等。在研究工作中,开拓性和探索性地进行了多光谱遥感数据的矿化蚀变信息提取建模、地球化学信息的模式识别、多源地学数据的融合处理等方面的技术方法研究,取得了具有创新性的成果。本文的研究成果主要有如下几个方面:1、通过对地质、遥感、地球化学和地球物理等多源地学数据的找矿信息提取和综合处理技术研究,形成了一套多元信息应用于矿产资源潜力评价和综合找矿预测的合理有效的技术方法组合;2、依据遥感图像的成像机理、多光谱遥感数据的波段特征,以及地物的地面光谱特征,结合地质矿产勘查基础理论,采用多元数据分析、图像增强处理、图像分割(分类)、图像掩膜等技术,研制出一系列新的多光谱遥感信息建模和提取方法;3、依据地质演化过程中的地球化学元素异常组合关系,提出并开发出用于化探数据综合异常信息挖掘的模式识别方法;4、根据多源地学数据的信息特征和应用目的,通过它们之间的成因和空间关系研究,总结和设计出不同信息抽象层次的多源数据融合处理的技术方法;5、基于GIS技术,采用本文研究成果,对东天山地区进行了矿产资源潜力的综合评价。更多还原

【Abstract】 Current mineral resource prospecting has already entered the era of high-techapplication from traditional prospecting method,and the compositive prospectingmeans,which combine remote sensing with geology,geophysics and geochemistry,has become modern ore-finding technique’s mainstream.Currently,contradictionbetween the supply and demand of mineral resources is increasing,but the using rateof geological data resources is still lower,and the data processing ability is relativelag in our country.Therefore,this study undoubtedly has an important practical andtheoretic meaning to advance prospecting technique level,improve ore-finding effectand efficiency.This paper has studied the technique of information processing,modeling,extracting and synthetical excavating of regional geological and mineral data,multispectral remote sensing data ( TM,ETM,ASTER ),as well as regionalgeochemistry,aeromagnetics and gravitation data etc.,which applied an effectivemeans and a new kind of information for mineral resource potential evaluation of themineralization regions.The research contents in this paper includes:the remotesensing and ground observation data’s transforming,conforming,integrating andore-finding information extracting technique,the multivariate geological data’ssynthetical processing technique and the multivariate geological information’sapplication technique etc..In this research work,innovative and tentative studiesabout the mineralized alteration information extracting modeling of multispectralremote sensing data,geochemical information’s pattern recognition and multivariategeological data’s fusion processing technique have been made,and some innovativeresults have been achieved.The following are the primary research results in thispaper:1.By studying the ore-finding information extracting and synthetical processingtechnique of multivariate geological data such as geology,remote sensing,geochemistry and geophysics data,a set of combined techniques which can rationallyand effectively apply multivariate information to mineral resource potential evaluationand predicting are formed;2.Basing on remote sensing image’s forming mechanism,multispectral remotesensing data’s spectral characters,ground spectral characters of ground objects and thefoundation theories of mineral resources prospecting,a series of new methods ofmultispectral remote sensing information modeling and extracting are developed byusing multivariate data analyse,image enhancing,image division or classification andimage mask processing etc.;3.According to geochemical elements’ combination relation formed in the courseof geological evolvement,the pattern recognition method,which can be applied toexcavate the synthetical information of geochemical data,is brought forward anddeveloped;4.According to multivariate geological data’s information characters and appliedpurpose,the technique methods of multivariate data’s fusion processing in variousabstract levels are summarized and designed by studying their spatial and causalrelations;5.Basing on GIS technique and the research results in this paper,the applicationof mineral resources potential evaluation is made in east TianShan area.更多还原