摘要
城市下垫面包含多种不同年代、材料和成分的人工建筑物,其光谱多样性远远超过自然环境.利用高光谱遥感数据的丰富光谱信息,可以弥补传统遥感数据源(如航片、多光谱遥感数据等)在区分城市地物所需光谱分辨率等相关信息上的不足.从光谱分析与光谱匹配技术出发对城市地物和人工目标进行精细分类,可提供城市规划、环境监测、城市变迁乃至相关社会经济等方面的信息.本研究基于Hyperion高光谱遥感数据,以广州市区为试验区,尝试用手动提取终端像元,首先对水体与植被做分层掩膜,尽可能消除其在求取参考波谱与影像波谱角度过程中的影响;继而参考约翰-霍普金斯大学提供的标准光谱库的人工建筑物波谱,对光谱角度制图方法所产生的规则影像进行密度分割,获取与参考光谱夹角最小的端元准确位置,再通过高空间分辨率影像Quick B ird数据对其进行准实地验证,尽可能提"纯"并获取相应地物影像端元;最后,应用线性光谱分解模型提取出广州市区地表物质的丰度,由丰度图设定阈值生成地物分类图.结果表明:星载高光谱数据可识别出都市人工地物中的水泥混凝土、铺路混凝土、粘土瓦屋顶、较老建筑屋顶、裸土、高反射率未知物(玻璃、金属等)、低反射率未知物(阴影)、林地、草地与水体等,其总体精度为76.2099%,Kappa系数为0.7258.
Diverse spectra of material in urban is more complicated than natural environment＇ s, especially for various decades, difference of materials and its composition. Conventional multispectral urban land resolution. data associated with statistical data-driven methods are ineffective for extracting cover information in detail due to the insufficiency on spectral and Through targets of urban can about city planning spatial spectral analysis and spectral matching technology, land cover and artificial be classified discriminatingly. Results of classification provide information , environmental monitor, and variance of city. As the first spacebome sensor, NASA＇s EO-1 Hyperion now provide images in 242 spectral ands in the 400-2500 nm range compared to multispectral sensors such as Landsat ETM ＋, which provides only 6 broadbands over this wavelength region. The research in this study obtains the reflectance of land cover material from a series of preprocessing, geometric correction, radiance calibration, broad selection, and atmospheric correction via FLAASH, which can remove ＇ smile＇ effect. Spectra of field measurements can hardly be applied in classification directly because of major affects of several geographical aspects, such as topography, climate, etc. Particularly On the As the hotspet , the existence of mixed pixels results in the less widespread field measurements. other hand, end members in image can easily classify or identify the hyperspectral data. precondition of understanding hyperspectral data, end member extraction is always the hotspot. Existing methods are mainly non-empirical, semi-auto or auto extraction of end members. In this research, a pixel-based spectral analysis algorithm, spectral angle mapper （SAM）, is adopted to extract end members by matching image spectra to experimental spectra （by John Hopkins university spectral library）. Before application of SAM, image should be masked by water and vegetation to eliminate their affection in follow-up procedure. Density slice is applied in the rule images which is half-results of the SAM, in order to detect the end members in least angle. Then locate six end members, validate them by means of comparison of hyper spatial image data, Quick Bird, photographed at the near time. In the end, abundance of surface material in Guangzhou city is decomposed by linear spectral mixture model（LSMM） approach , and surface material map is derived from setting range for abundance map. The results confirm the ability of hyperspectral in urban material mapping, and can clearly classify the material in urban area with the total precision 76. 2099%, Kappa efficiency 0. 7258, as follow： cement concrete, paving concrete, clay tile roof, outdated building, bare soil, high reflectivity material（ e. g. glass, new metal）, low reflectivity material （ e. g. shadow）, forest, grass, and water.
出处
《应用基础与工程科学学报》
EI
CSCD
2009年第2期206-218,共13页
Journal of Basic Science and Engineering
基金
教育部科技研究重点项目(206107)
广州市教育局科技项目(08C026)
关键词
星载高光谱数据
地物识别
线性混合光谱分解
广州
spaceborne hyperspectral
material identify
linear spectral unmixing
Guangzhou
