遥感多光谱成矿信息提取与成矿预测

【作者】 张建国；

【导师】 胡祥昭； 朱谷昌；

【作者基本信息】 中南大学 ， 构造地质学， 2005， 硕士

【摘要】 遥感成矿信息是指遥感图像中反映成矿地质背景、控矿因素和岩石蚀变等与矿化有关的地质现象的遥感信息。笔者选择青海省督冷沟地区作为研究区,利用ETM卫星图像数据,进行与金属矿化有关的遥感成矿信息的提取方法研究及其在找矿中的应用研究。 研究工作中,首先对图像进行包括大气校正、几何校正和地理配准等的数据预处理;根据ETM图像辐值特征统计分析,选择彩色合成的最佳波段组合,并与ETM全色波段图像进行融合处理,获得研究区基础地质—地貌图像,并对该图像进行地质解译和分析,获取地层—岩石单元和断裂构造遥感信息。然后对典型蚀变矿物进行光谱分析,研究各类不同蚀变岩及围岩的光谱特征,以及它们在对应ETM图像各波段(ETM1～7)上的光谱特征及其光谱差异性,据此确立基于光谱特征的矿化蚀变遥感信息分类提取和增强提取的波段组合方案。然后对ETM图像进行掩模处理,分别采用主成分分析、比值处理等数据处理方法,定量提取矿化蚀变遥感信息,并采用IHS变换提取成矿信息,将其与前面所提的异常信息进行对比分析,进行效果评价,在此基础上,分析该区的成矿地质背景条件。最后,基于GIS技术,对获得的各遥感找矿信息以及已有的地理、地质矿产、物化探等数据进行系统集成、综合处理和分析,确定找矿标志,作出成矿预测;共圈定找矿靶区5处.经野外检查和验证,预测效果良好。更多还原

【Abstract】 Remote sensing mineralized information is remote sensing information which reflects geologic phenomenon related to mineralization such as mineralizing geological background, mineral control factors and alterative rock. The author selected Dulenggou area in Qinghai province as research district, uses ETM satellite image data to study the extraction methods of remote sensing mineralized information and its application on mineral exploration.The first step of the research was data pretreatment to image which includes atmosphere emendation, geometry emendation and geography registration and so on. Then making an image of basic geology-physiognomy by selecting best bands combination to color composition according to statistical analysis to value feature of the ETM image, and fusing with ETM’s pan band to obtained remote sensing information of stratum-rock unit and fault structure by geologic explanation and analysis. Then making spectral analysis to representative alterative mineral and studying the spectral feature of each kind of alterative rock and wall rock along with their spectrum feature correspond to every band of ETM image (ETM 1-7) and their spectral difference. Hereby, establishing method of band combination for classifying extract and enhancement extract to remote sensing mineralized information based on spectrum vector. Then masking the ETM image and extracting remote sensing mineralized information by principal component analysis and ratio processing. To compare with abnormity information which extracted in front, and judge its effect, adopted IHS transform to extract mineralized information. Then analyzing background condition of mineral geology. At last, based on GIS technology, making system integration and colligate process and analysis to data such as kinds of remote sensing mineralized information and geography, geologic mineral resources, geophysical prospecting and geochemistry prospecting and so on, then making metallogenic prediction. Enclosed 5 prospecting areas which take good effect by field examination.更多还原