成像光谱矿物填图精度兼与矿物丰度关系初探

成像光谱遥感技术在岩石矿物识别和矿物填图方面发挥着越来越大作用,而成像光谱矿物填图的精度问题一直是研究的焦点。文章利用高光谱数据HyMap对新疆东天山地区的矿物填图结果,通过引入矿物填绘分布强度的概念,探索性的分析了成像光谱矿物填图与填绘矿物丰度的关系。     

改进相似性测度的高光谱影像FCM聚类分析

高光谱遥感将反映目标辐射属性的光谱信息与反映目标空间几何关系的图像信息有机地结合在一起,能够实现地面目标的精细分类识别。FCM是一种有效的聚类算法,但存在相似性测度模型单一、分类精度的提高受到限制等问题。文中结合高光谱影像的技术特点,综合考虑光谱曲线的形状、地物辐射亮度及其权重,提出可以更好描述光谱向量之间的相似性的距离测度,并将其引入到FCM聚类模型中。聚类分析试验结果表明：通过改进和优化相似性测度的FCM,可以显著提高高光谱影像聚类精度。     

CHRIS／PROBA高光谱数据的预处理

针对CHRIS／PROBA高光谱遥感数据的特点,对CHRIS影像进行去条带处理、辐射校正、大气校正等预处理,获得较好质量的影像,为影像的进一步分析和实际应用提供保障。     

端元光谱自动提取的总体最小二乘迭代分解

根据总体最小二乘(total least squares, TLS)模型理论,提出了一种影像端元光谱可受噪声污染的混合光谱线性扩展模型,并实现了该模型的端元光谱自动迭代提取以及混合像元的限定性分解.实验结果表明,扩展的混合像元分解模型明显优于传统的最小二乘分解模型,总体精度大约提高了10%～20%.     

基于MRMRF的多光谱纹理影像分割方法

提出了一种新的基于多分辨率马尔可夫场(multiresolution Markov random field, MRMRF)模型的小波域多光谱纹理影像分割方法.该方法采用可变的权重参数连接小波分解的多尺度特征场和标记场,通过直接映射,将上一尺度的分割结果作为下一尺度的初始结果,最细尺度上的分割结果作为算法的最终分割结果.实验表明,本文算法能够降低多光谱纹理图像分割的错误率.     

新鲜煤和氧化煤自燃倾向性的FTIR对比分析

利用傅立叶变换红外光谱（FTIR）分析技术,对乌达矿区低温氧化和新鲜煤样进行了红外光谱测试,通过两类煤样的特征官能团吸光度的对比分析．揭示了煤在氧化前后特征官能团的变化规律：氧化后的煤样中脂肪族和芳烃CH减少,酸酐C=O从无到有,芳烃骨架C=C基本不变,进而研究了其对煤自燃倾向性的影响。结果表明脂肪族和羟基含量越高,煤自燃倾向性越大。     

光谱快速预选筛分算法研究

利用一组标准正交化的参考向量得到一个可能匹配区域,排除了绝大部分不可能匹配的样本,从而大大提高光谱预选筛分的速度;同时介绍了一种平衡速度与精度的编码替代算法.实验证明:压缩影像的目视效果较好,未产生边缘模糊及条带现象;采用自动目标生成方法(ATGP)从样本集中提取端元并进行光谱角度制图,其效果令人满意.     

Characterization of cristobalite in ash and dome rock from the Soufriere Hills volcano, Montserrat： Implications for respiratory health hazards

A link between the inhalation of respirable silicas （SiO2） and respiratory diseases such as silicosis is widely recognized. Ash from dome collapse eruptions on Montserrat has been found to contain high levels （〉20%） of silicas in the form of cristobalite, tridymite, quartz and/or amorphous silica. The toxicity of these silica polymorphs varies widely. Cristobalite and quartz （tridymite less well established） are viewed as carcinogenic to humans whereas amorphous silica generally shows a reduced biological response. In assessing the potential health effects of volcanic ash particulates it is vital to determine the types and concentration of silicas as well as their size （respirable fraction）, shape and surface properties. The aim of this study is to develop methods to assess potentially toxic respirable airborne silicas in the dome collapse ash （applicable to a range of ash types） and to develop a model to predict the levels and types of respirable silicas from future eruptions. The model is being developed by comparing dome rock with related ash from a series of previous eruption events. Mineralogical assessment using conventional scanning X-ray diffraction （XRD） was hampered by difficulties in differentiating characteristic peaks for cristobalite and tridymite in complex multi-component ash samples （containing high levels of plagioclase）. These difficulties have been largely overcome using an Enraf-Nonius PDS120 diffractometer with curved （120 degrees 20） position sensitive detector （PSD）. The determination of size, shape and elemental characteristics of ash particulate and dome rock samples has been carried out using automated analytical scanning electron microscopy. The quantification of mineral proportions using PSD-XRD was highly successful with an accuracy of 1 to 2 wt%. However, the determination of phase proportions using automated analytical SEM was problematic due to scattering effects and the multiphase nature of many of the particles.     

Origin and three-dimensional fluorescence spectrum of the chromophoric dissolved organic matter in Yangtze estuary, East China

Dissolved organic matter （DOM） is an important chemical component in natural water. Chromophoric dissolved organic matter （CDOM）, a fraction of optical properties, plays art important role in the biogeochemical cycle of nutrients in aquatic environment. People realized that DOM cycle is crucial in the global carbon and nitrogen flux, and also is inherently related to nutrients and trace metal elements. Therefore, CDOM was concerned by scientists in global oceanography and limnology fields. Water samples were collected from three sections （North Channel, South Channel and Zhuyuan） of the Yangtze （Changjiang River） estuary in March 2006 Three-dimensional excitation emission matrix （3-DEEM） fluorescence spectra were analyzed for those filtrates through Whatman GF/F filters. Dissolved organic carbon （DOC） was also measured by TOC analyzer. The tidal variety was also taken into account. The 3-D EEM fluorescence scans suggested the fluorescence characteristics of humic acid （Ex=332-344 nm, Em=439-451 nm） and fulvic acid （Ex=250-254 nm, Em=472-478 nm） were obvious, and the fluorescence group of protein-like and tyrosine （Ex=230 nm, Em=283 nm） was also found. They are mainly composed of CDOM in the Yangtze estuary. Further data analysis, especially the fluorescence index （f 450/500）, showed that terrestrial signal was rather strong （1.41-1.65） in the surface water, however, some terrestrial CDOM signals of bottom water showed excursions （1.28-1.39）. On the other hand, anthropogenic sign was impressed in the waters of Zhuyuan, which is one of the main drain outlets of Shanghai Metropolis. DOC concentrations ranged from 2.2 mg/L to 3.4 mg/L in Zhuyuan and South Channel, and from 2.0 mg/L to 2.4 mg/L in North Channel. The tide effect played a role in the composition of the CDOM measured by 3-D fluorescence scan technology.