【作者】 阮建武；

【导师】 邢立新； 潘军；

【作者基本信息】 吉林大学 ， 地球探测与信息技术， 2004， 硕士

【摘要】 遥感技术是以电磁波辐射为表现形式，通过各种有效的手段来收集、处理、分析和提取所需要的特征，达到识别所研究对象的存在、状况和动态的目的。作为一种采集地球数据及其变化信息的重要技术手段，遥感在许多国家的政府部门、科研单位和公司得到了广泛的应用。现代遥感技术最重要的形式就是通过各种遥感器，收集地面物体在某一电磁波辐射范围内所反映的图像，通过对获取图像的分析、研究，达到识别物体的目的。可见，任何遥感图像都是地物电磁波谱特性的客观记录，遥感图像的分析解译与应用处理过程，就是一个从记录电磁波谱辐射能量的图像反推地物目标属性类别及其时空分布变化的过程。因此，研究地球表面各种物体电磁波辐射特性是遥感技术的一项极重要的基础，也是推动遥感技术发展必不可少的工作。遥感科学应用中主要研究地物电磁波谱的反射、发射和微波特性，主要利用的波段包括0.36～2.5μm的反射光谱段，3～5μm的反射发射光谱段，8～14μm的发射光谱段和大于1cm的某些微波谱段。目前，遥感研究地物波谱的内容主要有：①地物波谱测量与多光谱遥感最佳波段选择；②遥感信息模型研究；③热红外和微波波谱特性研究；④地物波谱特性的物理数学模式研究；⑤地物波谱在成像光谱数据处理中的应用。随着遥感技术定性向定量化方向的发展，数据源的波谱分辨率和空间分辨率的提高，如太空成像公司(Space Imaging)开发的IKONOS卫星带有高达1m空间分辨率的全色波段，早期使用的多波段式低光谱分辨率的光谱仪已逐渐被具有高光谱分辨率的光谱仪所替代，而由此所引起的数据量日益庞大问题对地物光谱数据的收集、积累和管理提出了更新、更高的要求。建立地物波谱数据库，运用先进的计算机技术来保存、管理和分析这些信息，一方面可以改进与提高对地物波谱信息的处理、分析能力与效率，加强遥感基础理论的研究，另一方面为地物波谱数据在计算机图像处理、自动判读的应用研究工作提供有效手段，同时为遥感技术及其应用人员提供方便、快速的查询系统及灵活、多样的数据处理<WP=65>功能，促进遥感技术和应用研究的深入发展。随着遥感应用的深入，遥感信息与地物相互作用的研究有了进一步发展，特别是成像光谱仪的应用，不仅显示了地物波谱特性研究的重要性，而且也推动了这一领域的研究。因为它可以获得图谱合一的信息，可以直接将地物波谱特性和遥感图像结合在一起，在图像分析和应用方面都取得了很好的效果。本论文在研究地物波谱数据特性基础上，采用Visual Basic 6.0设计并实现波谱数据库的编辑、查询、分析等功能，研究地物波谱特征与遥感信息之间的关系，为识别地物提供理论依据。数据库的光谱分析处理方法主要有光谱特征提取、光谱匹配识别和光谱角度制图法。同时，以距离系数和夹角余弦作为匹配识别的相似性度量参数，分别对非成像光谱数据和多光谱数据进行实例分类研究，其中，非成像光谱数据选用JHU、USGS和JPL波谱库中的部分数据以及满洲里地区野外实测光谱数据；多光谱数据选用满洲里地区Landsat/TM遥感图像。另外，对TM数据进行分类研究时，由于原始遥感数据中包含的畸变信息是影响分类应用效果的重要因素，因此必须采用适当的方法对图像进行预处理，消除或部分消除图像畸变信息，本文利用PCI 软件的ATCOR2模块对原始数据进行了大气校正。对非成像光谱数据和多光谱成像数据进行的匹配识别，取得了较好的效果，为进一步的信息提取提供了参考。更多还原

【Abstract】 Remote Sensing is one set of comprehensive technology, which is based on the electromagnetic wave, through various methods to collect, process, analyze and extract the needed feature so as to recognize the existence, conditions and dynamic situations of the studied objects. As a kind of important technology used to collect geo-science data and the changing information of the Earth, Remote Sensing has been widely used in many countries and their various institutions.The prime form of contemporary Remote Sensing technology is as follows: through all kinds of sensors, collecting the imaging data within a specific electromagnetic spectrum and through analyzing and studying the acquired images so as to recognize the objects interested. That is to say, any remotely sensed image is objective record of the electromagnetic characteristic of ground objects. The process of interpretation and application of remotely sensed images is to retrieve the attributes and space-time dynamics of ground objects from the images that carry electromagnetic radiation information. Therefore, it is a key foundation of Remote Sensing technology to study the electromagnetic radiation characteristic of various ground objects, which in the meantime is an indispensable work to impel the progress of Remote Sensing technology.In the application of Remote Sensing, the refection, emission and microwave characteristic are mainly concerned and the primary spectral range includes reflection band: 0.36～2.5μm, reflection and emission band: 3～5μm, emission band: 8～14μm, and some microwave bands: >1cm. Nowadays, the contents of the study of ground objects spectral characteristic mainly include (1) the measurement of ground objects’ spectra and the selection of the optimal window of multi-spectral Remote Sensing; (2) the study of remote sensing information models; (3) the study of thermal and <WP=67>microwave spectral characteristics; (4) the study of physical and mathematical patterns of ground objects’ spectral characteristics; (5) the application of ground objects’ spectra in the processing of imaging spectral data.Along with the development of Remote Sensing from the quality-oriented to the quantity-oriented and the increase in the spectral and spatial resolution of data source, such as 1-meter resolution of panchromatic band mounted on the Satellite IKONOS developed by Space Imaging Company, the relatively low resolution multi-spectral sensors have been gradually replaced by the high resolution counterparts. Nevertheless, the consequential problems of ever-increasing data amount engender new demands with regard to the collection, accumulation, and management of ground objects’ spectra. Building spectra database of ground objects and using advanced computer technology to save, manage and analyze these data, on the one hand, can improve the capability and efficiency of the processing and analyzing of spectral information and strengthen the foundational research of Remote Sensing, and on the other hand can provide effective means for the application personnel. In the meantime, they provide both research and application personnel with convenient and rapid query system and flexible and adverse data processing function and promote the further development of research and application Remote Sensing.Along with the comprehensive application of Remote Sensing, the study of the interaction between Remote Sensing information and ground objects has further development. Especially, the application of imaging spectrometer not only reveals the importance of the study of spectral characteristic of ground objects but also promote the study in this domain. The spectrometer has achieved very good effects in image analyzing and application. The thesis, on the base of studying spectral characteristics of ground objects, uses Visual Basic 6.0 to design and implement database functions such as edition, query and analysis and studies the relationship between ground objects’ spectral <WP=68>characteristics and remotely sensed information, which provide theore更多还原