基于特征谱带的高光谱遥感矿物谱系识别

文中初步提出并建立矿物识别谱系的分层识别技术以及部分离子、矿物的光谱识别规则。矿物光谱包含一系列特征吸收谱带。这些特征谱带在不同的矿物中具有较稳定的波长位置和较稳定的独特波形 ,能够指示离子类矿物、矿物的存在 ,是利用高光谱进行矿物识别的基础。文中分析了部分在自然界广泛存在的一些阴阳离子 (如CO2 -3 ,Fe2 + ,Fe3 + ,Al3 + ,Mn2 + ,Al—OH和Mg—OH等 )的可识别特征光谱 ,总结部分离子类矿物或矿物识别规则 ,并以美国Cuprite地区的AVIRIS数据进行矿物识别试验 ,利用矿物谱系识别技术从离子类矿物到单矿物再到更精细类矿物逐层开展识别 ,取得了较好的效果 ,初步实现利用宏观的手段 (遥感技术 )进行微观 (矿物 )的探测。

SPECTRAL IDENTIFICATION TREE(SIT)FOR MINERAL EXTRACTION BASED ON SPECTRAL CHARACTERISTICS OF MINERALS

The Spectral Identification Tree (SIT) and the spectral identification rules based on the profiles of the characteristic spectra of rocks and minerals was designed and developed to extract minerals directly using hyperspectral remote sensing data. Minerals have some special diagnostic absorbing features that have relatively steady wavelength positions and widths, especially the unique steady going spectral profiles. These chief spectral parameters, such as the diagnostic absorbing feature and the feature association zone can be used as the indicators of ions and minerals. The spectral absorption features of different ions or bonds such as CO 2- 3, Fe 2+ , Fe 3+ , Mn 2+ , Al-OH and Mg-OH were studied. Especially, some main and absorption features corresponding to various types of ions or minerals were analyzed and summarized. Some Spectral Identification Trees (SIT) from various ions to minerals was established based on all of above spectral absorption features. Various spectral absorption features were combined to extract and discriminate the different types of minerals in various levels by SIT technique. The information of ions and bonds, such as CO 2- 3, Fe 2+ , Fe 3+ , Mn 2+ , Al-OH and Mg-OH, and of some minerals, such as alunite, montmorillonite, muscovite, halloysite, smectkaolinite and illite were identified and extracted effectively by SIT technique using AVIRIS data from Cuprite, Nevada of USA, which was chosen as a test area.