基于遥感影像的新疆西天山地区构造及岩性信息提取研究
发布时间：2018-11-15 14:09
【摘要】：遥感地质是有效利用最先进的遥感技术研究地质变化规律的学科,是区域化地质填图和矿产勘查的有效方法之一。它是在各种地上物质对电磁辐射有不同反应的基础上,主要用遥感的方法研究构造、岩性、蚀变等信息的解译,并结合其他相关资料,以达到分析、判断、识别一定地区内地质规律的目的。区域性地质构造解译,特别是环形构造与线性构造的解译是地质勘查前期的主要参考;区域性岩性信息的提取是地质勘查中后期的必要内容。本文以新疆西天山地区的查岗诺尔--智博--敦德--备战一带为研究区,以Landsat-7、Landsat-8和Google Earth等多源遥感数据为依托,整合野外采集GPS点数据及实测岩石光谱数据,结合遥感地质相关理论,综合其它物化探等信息,重点研究了构造信息和岩性信息的提取,并进行了相关信息对应分析及找矿远景区预测。主要研究成果如下:(1)对遥感图像进行预处理。通过图像的几何校正,基于Flaash变换的大气纠正,基于小波HIS变换的图像融合,图像的镶嵌和裁剪,基于光谱变换的图像增强,得到更适合进行构造和岩性信息提取的遥感影像。(2)构造信息解译。在预处理后的ETM+影像基础上,借助ArcScene软件平台生成了模拟三维遥感影像图,结合构造地质学理论,野外调研对构造的认识,研究区地理背景及其他相关地质资料,解译出了多条大中型的线性构造和多个典型的环形构造。(3)岩石光谱信息与影像灰度值对应分析。通过成像光谱仪通道重采样等方法对研究区几种典型岩石光谱进行处理分析。在预处理后的OLI影像基础上,提取研究区这几种典型岩石不同波段的灰度值,并通过影像灰度值的平均、归一化等方法进行处理分析。同时运用了一种特征信息提取的方法(对应分析法),对岩石光谱和影像灰度值之间的相互关系进行了探究。最后运用SPSS软件对重采样后的光谱曲线和归一化后的影像灰度值曲线进行线性回归分析,并拟合出了最佳回归模型。(4)面向对象的岩性解译。在岩石光谱与影像灰度值对应分析的基础上,结合岩石的纹理、构造、颜色、结构、成分、胶结物及胶结类型等信息,对研究区进行区域岩性分类,并与对应地质图比对验证分析。(5)成矿远景区预测。结合构造、岩性、物探、化探、相关地质背景、Google earth影像等信息,对研究区进行综合异常标志分析并预测成矿远景区。
[Abstract]:Remote sensing geology is one of the most effective methods of regional geological mapping and mineral exploration, which can effectively use the most advanced remote sensing technology to study the law of geological change. On the basis of the different reactions of various aboveground materials to electromagnetic radiation, it mainly uses remote sensing methods to study the interpretation of structural, lithologic, alteration and other information, and combines other relevant data to achieve analysis and judgment. The purpose of identifying geological laws in a certain area. The interpretation of regional geological structure, especially the interpretation of ring structure and linear structure, is the main reference in the early stage of geological exploration, and the extraction of regional lithologic information is the necessary content in the middle and late period of geological exploration. In this paper, based on the multi-source remote sensing data such as Landsat-7,Landsat-8, Google Earth and other remote sensing data, this paper integrates the field GPS data and the measured rock spectral data based on the study area of Chagannul, Zhibo and Dunde-Preparedness area in the West Tianshan area of Xinjiang, and based on the multi-source remote sensing data such as Landsat-7,Landsat-8 and Google Earth. Combined with the theory of remote sensing geology and other geophysical and geochemical exploration information, the extraction of structure information and lithologic information is studied, and the corresponding analysis of relevant information and the prediction of prospecting area are carried out. The main research results are as follows: (1) preprocessing remote sensing image. Through image geometric correction, atmospheric correction based on Flaash transform, image fusion based on wavelet HIS transform, image mosaic and clipping, image enhancement based on spectral transformation. Remote sensing images which are more suitable for structural and lithologic information extraction are obtained. (2) structural information interpretation. On the basis of preprocessed ETM images, the simulated 3D remote sensing images are generated with the help of ArcScene software platform. Combined with the theory of tectonic geology, the understanding of structure, the geographical background of the study area and other related geological data, Several large and medium-sized linear structures and several typical ring structures are derived. (3) the corresponding analysis of the spectral information of rock and the gray value of image. Several typical rock spectra in the study area were processed and analyzed by imaging spectrometer channel resampling. Based on the pre-processed OLI images, the gray values of these typical rock bands in the study area are extracted, and processed and analyzed by means of the average and normalized gray values of the images. At the same time, a method of feature information extraction (correspondence analysis) is used to explore the relationship between the spectrum of rock and the gray value of image. Finally, the linear regression analysis of the spectral curve and the normalized image gray value curve after resampling is carried out by using SPSS software, and the best regression model is fitted. (4) Object-Oriented lithology interpretation. On the basis of the correspondence analysis of rock spectrum and image gray value, combined with the information of texture, structure, color, structure, composition, cementation and cementation type of rock, this paper classifies the regional lithology of the study area. And compared with the corresponding geological map verification analysis. (5) metallogenic area prediction. Combined with the information of structure, lithology, geophysical exploration, geochemical exploration and related geological background, Google earth image, the study area is analyzed with comprehensive anomaly markers and the metallogenic area is predicted.
【学位授予单位】：长安大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P627
本文编号：2333519
[Abstract]:Remote sensing geology is one of the most effective methods of regional geological mapping and mineral exploration, which can effectively use the most advanced remote sensing technology to study the law of geological change. On the basis of the different reactions of various aboveground materials to electromagnetic radiation, it mainly uses remote sensing methods to study the interpretation of structural, lithologic, alteration and other information, and combines other relevant data to achieve analysis and judgment. The purpose of identifying geological laws in a certain area. The interpretation of regional geological structure, especially the interpretation of ring structure and linear structure, is the main reference in the early stage of geological exploration, and the extraction of regional lithologic information is the necessary content in the middle and late period of geological exploration. In this paper, based on the multi-source remote sensing data such as Landsat-7,Landsat-8, Google Earth and other remote sensing data, this paper integrates the field GPS data and the measured rock spectral data based on the study area of Chagannul, Zhibo and Dunde-Preparedness area in the West Tianshan area of Xinjiang, and based on the multi-source remote sensing data such as Landsat-7,Landsat-8 and Google Earth. Combined with the theory of remote sensing geology and other geophysical and geochemical exploration information, the extraction of structure information and lithologic information is studied, and the corresponding analysis of relevant information and the prediction of prospecting area are carried out. The main research results are as follows: (1) preprocessing remote sensing image. Through image geometric correction, atmospheric correction based on Flaash transform, image fusion based on wavelet HIS transform, image mosaic and clipping, image enhancement based on spectral transformation. Remote sensing images which are more suitable for structural and lithologic information extraction are obtained. (2) structural information interpretation. On the basis of preprocessed ETM images, the simulated 3D remote sensing images are generated with the help of ArcScene software platform. Combined with the theory of tectonic geology, the understanding of structure, the geographical background of the study area and other related geological data, Several large and medium-sized linear structures and several typical ring structures are derived. (3) the corresponding analysis of the spectral information of rock and the gray value of image. Several typical rock spectra in the study area were processed and analyzed by imaging spectrometer channel resampling. Based on the pre-processed OLI images, the gray values of these typical rock bands in the study area are extracted, and processed and analyzed by means of the average and normalized gray values of the images. At the same time, a method of feature information extraction (correspondence analysis) is used to explore the relationship between the spectrum of rock and the gray value of image. Finally, the linear regression analysis of the spectral curve and the normalized image gray value curve after resampling is carried out by using SPSS software, and the best regression model is fitted. (4) Object-Oriented lithology interpretation. On the basis of the correspondence analysis of rock spectrum and image gray value, combined with the information of texture, structure, color, structure, composition, cementation and cementation type of rock, this paper classifies the regional lithology of the study area. And compared with the corresponding geological map verification analysis. (5) metallogenic area prediction. Combined with the information of structure, lithology, geophysical exploration, geochemical exploration and related geological background, Google earth image, the study area is analyzed with comprehensive anomaly markers and the metallogenic area is predicted.
【学位授予单位】：长安大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：P627
【引证文献】
相关会议论文 前1条
1 李美娟;陈国宏;陈衍泰;;综合评价中指标标准化方法研究[A];2004年中国管理科学学术会议论文集[C];2004年
，本文编号：2333519
本文链接：https://www.wllwen.com/kejilunwen/diqiudizhi/2333519.html
