利用卷积神经网络的高光谱图像分类     

赵漫丹  任治全  吴高昌  郝向阳 《测绘科学技术学报》2017,(5):501-507

针对高光谱图像分类中对光谱信息利用不足的问题,提出一种基于卷积神经网络在光谱域开展的分类算法。该算法通过构建五层网络结构,逐像素对光谱信息开展分析,将全光谱段集合作为输入,利用神经网络展开代价函数值的计算,实现对光谱特征的提取与分类。实验中采用三组高光谱遥感影像数据进行对比分析,以India Pines数据集为例,提出的基于卷积神经网络的分类方法的分类正确率达到90.16%,比RBF-SVM方法高出2.56%,相比三种传统的深度学习方法高出1%~3%,训练速度也较为理想。实验结果表明,本文所提出的算法充分利用了高光谱图像中逐像素点的光谱域信息,能够有效提高分类正确率。与传统学习算法相比,在较少训练样本的情况下,更能发挥其良好的分类性能。  相似文献   

Remote detection of fluid-related diagenetic mineralogical variations in the Wingate Sandstone at different spatial and spectral resolutions     

《International Journal of Applied Earth Observation and Geoinformation》2016

Well-exposed eolian units of the Jurassic system on the Colorado Plateau including the Wingate Sandstone, show prominent color variations throughout southeastern Utah due to diagenetic changes that include precipitation and/or removal of iron oxide, clay, and carbonate cement. Spatially variable characteristic diagenetic changes suggest fluid-rock interactions through the sandstone. Distinctive spectral signatures of diagenetic minerals can be used to map diagenetic mineral variability and possibly fluid-flow pathways. The main objective of this work was to identify characteristic diagenetic minerals, and map their spatial variability from regional to outcrop scale in Wingate Sandstone exposures of Lisbon Valley, Utah. Laboratory reflectance spectroscopy analysis of the samples facilitated identification of diagnostic spectral characteristics of the common diagenetic minerals and their relative abundances between altered and unaltered Wingate Sandstone. Comparison of reflectance spectroscopy with satellite, airborne, and ground-based imaging spectroscopy data provided a method for mapping and evaluating spatial variations of diagenetic minerals. The Feature-oriented Principal Component Selection method was used on Advanced Spaceborne Thermal Emission and Reflection Radiometer data so as to map common mineral groups throughout the broader Wingate Sandstone exposure in the area. The Minimum Noise Fraction and Spectral Angle Mapper methods were applied on airborne HyMap and ground-based hyperspectral imaging data to identify and map mineralogical changes. The satellite and airborne data showed that out of 25.55 km² total exposure of Wingate Sandstone in Lisbon Valley, unaltered sandstone cover 12.55 km², and altered sandstone cover 8.90 km² in the northwest flank and 5.09 km² in the southern flank of the anticline. The ground-based hyperspectral data demonstrated the ability to identify and map mineral assemblages with two-dimensional lateral continuity on near-vertical rock faces. The results showed that 39.71% of the scanned outcrop is bleached and 20.60% is unbleached while 6.33% remain unclassified, and 33.36% is masked-out as vegetation. The bleached and unbleached areas are alternating throughout the vertical face of the outcrop. The relative hematite abundance observed in the unbleached areas are somewhat symmetrical. This indicates fairly similar reaction intensities along the upper and lower reaction fronts observed in the vertical section. The distribution geometry and relative abundances of diagenetic minerals not only suggest multiple paths of fluid-flow in Wingate Sandstone but also provides some insight about relative direction of past fluid-flow.  相似文献   

高光谱影像空-谱协同嵌入的地物分类算法   总被引：4，自引：4，他引：0  

黄鸿  郑新磊 《测绘学报》2016,45(8):964-972

针对传统高光谱影像地物分类算法大多仅考虑光谱信息而忽略空间邻近像元间相关性的问题,提出了一种空-谱协同嵌入(SSCE)降维算法和空-谱协同最近邻(SSCNN)分类器。首先,定义一种空-谱协同距离,并将其应用于近邻选取和低维嵌入;然后,构建空-谱近邻关系图来保持数据中的流形结构,并在权值设置中增大空间近邻点的权重以增强数据间的聚集性,提取鉴别特征;最后使用SSCNN分类器对降维后的数据进行分类。利用PaviaU和Salinas高光谱数据集进行试验验证,结果表明,与传统的光谱分类算法相比,该算法能有效提高高光谱影像的地物分类精度。  相似文献   

线性/非线性光谱混合模型估算白刺灌丛植被覆盖度     

姬翠翠  贾永红  李晓松  王金英 《遥感学报》2016,20(6):1402-1412

及时监测干旱与半干旱区光合/非光合植被覆盖度时空变化,可以为指导荒漠化防治工程及植被衰退机制研究提供重要信息。本文以甘肃民勤典型植被白刺灌丛为研究对象,通过地面控制性光谱实验获取混合光谱、端元光谱与丰度信息,开展线性与非线性光谱混合模型(包括核函数非线性和双线性混合模型)估算光合和非光合植被覆盖度的对比研究,采用全限制最小二乘法进行模型解混,分别获取各样本数据中各类端元丰度及其精度信息,通过模型分解的均方根误差(RMSE)与地面验证精度确定用于光合和非光合植被覆盖度估算的最佳光谱混合模型,其中参考端元丰度采用神经网络(NNC)分类算法对数字影像进行分类获取。结果表明:(1)引入阴影端元的四端元模型相对于传统的三端元模型(光合/非光合植被与裸土)能有效提高光谱解混的精度,并提高光合和非光合植被覆盖度估算精度;(2)对白刺灌丛来说,光合植被、非光合植被、裸土及阴影间多重散射混合效应存在,但混合效应不够显著;考虑非线性参数的核函数非线性光谱混合模型表现略低于线性光谱混合模型,因此非线性光谱混合模型在估算白刺灌丛光合和非光合植被覆盖度时相对于线性光谱混合模型没有明显优势;(3)基于光合/非光合植被、裸土与阴影四端元的线性光谱混合模型可以实现白刺灌丛光合和非光合植被覆盖度的准确估算,光合植被覆盖度估算RMSE为0.11 77,非光合植被覆盖度估算RMSE为0.0835。  相似文献   

高光谱测量曲线中干扰信号的消除方法   总被引：1，自引：0，他引：1  

朱芸  魏广凯  郑秋生 《海洋科学进展》2002,20(4):76-78

现场获得的高光谱数据是多种信号源信息的叠加，其中包括因仪器噪音和环境干扰形成对测量数据目标信息的污染。在对光谱数据分析之前，需对其进行必要的处理，以消除非目标信息的干扰。结合部分现场高光谱数据介绍了几种滤波方法，并对应用效果进行了分析。  相似文献   

基于高光谱影像数据库的地理目标识别     

张立朝  孙婷  唐汉松 《海洋测绘》2007,27(1):51-54

高光谱技术的发展使得通过光谱匹配进行地理目标识别成为可能。围绕光谱分解、分类以及匹配算法进行研究，设计了一套规范实用的高光谱数据库系统，构建了高光谱影像样本库。提出了基于高光谱影像数据库进行地理目标识别的解决方案。  相似文献   

Imaging spectroscopy as a tool to study sediment characteristics on a tidal sandbank in the Westerschelde     

B. Deronde  P. Kempeneers  R.M. Forster 《Estuarine, Coastal and Shelf Science》2006,69(3-4):580

This paper focuses on the use of imaging spectroscopy for the mapping of sediment characteristics on a tidal sandbank in the Westerschelde, called the Molenplaat. On June 8, 2004, during low tide, a HyMap™ scanner recorded the Molenplaat at 4 m pixel resolution. The hyperspectral data were radiometrically calibrated, geometrically corrected, and atmospherically corrected to give apparent surface reflectance data. On the calibrated and corrected dataset a supervised binary classification was performed, based on linear discriminant analysis. Simultaneous to the flight, 25 sediment samples were collected in the field and analysed in the lab to define the median grain size, the water content, the total organic matter content and the chlorophyll-a concentration. These four parameters play a crucial role in sediment stability and macrofaunal habitat definition. Prior to the classification, a feature selection, based on sequential floating forward search (SFFS), was performed. For each of the four parameters two to three bands were retained for the classification. These bands were most frequently selected in the visible and near infrared parts of the spectrum, except for the organic matter content where also SWIR bands were used. The overall classification accuracy was highest for the water content (88%), the median grain size (88%) and the chlorophyll-a concentration (84%). The organic matter content, for which three instead of two classes were distinguished, scored somewhat lower but still reached 80%. The classifications were limited to a small number of classes in order to obtain reliable statistics with a small number of training samples. The spatial patterns in the classified images indicated that the four parameters under study are highly correlated. In most cases coarse sediment coincided with dry conditions, low organic matter and low concentrations of chlorophyll-a. The wet and muddy parts of the Molenplaat were in general characterised by a notably higher amount of organic matter and chlorophyll-a. The individual classification results for the median grain size, the water content and the chlorophyll-a concentration were combined to generate a sediment ecotope map. The presented study illustrates how airborne hyperspectral data can be used to achieve accurate classified maps of intertidal sediment ecotope types, applying feature selection and a binary classification approach.  相似文献   

Discrimination and classification of mangrove forests using EO-1 Hyperion data: a case study of Indian Sundarbans     

Tanumi Kumar  Abhishek Mandal  Dibyendu Dutta  R. Nagaraja  Vinay Kumar Dadhwal 《国际地球制图》2019,34(4):415-442

In remote sensing the identification accuracy of mangroves is greatly influenced by terrestrial vegetation. This paper deals with the use of specific vegetation indices for extracting mangrove forests using Earth Observing-1 Hyperion image over a portion of Indian Sundarbans, followed by classification of mangroves into floristic composition classes. Five vegetation indices (three new and two published), namely Mangrove Probability Vegetation Index, Normalized Difference Wetland Vegetation Index, Shortwave Infrared Absorption Index, Normalized Difference Infrared Index and Atmospherically Corrected Vegetation Index were used in decision tree algorithm to develop the mangrove mask. Then, three full-pixel classifiers, namely Minimum Distance, Spectral Angle Mapper and Support Vector Machine (SVM) were evaluated on the data within the mask. SVM performed better than the other two classifiers with an overall precision of 99.08%. The methodology presented here may be applied in different mangrove areas for producing community zonation maps at finer levels.  相似文献   

高光谱图像滚动引导递归滤波与地物分类     

崔宾阁  吴亚男  钟勇  钟利伟  路燕 《遥感学报》2019,23(3):431-442

高光谱图像类内光谱变化较大,"同物异谱"现象普遍存在。利用原始地物光谱特征进行分类精度较低而且分类结果图中存在"椒盐现象"。为了获得好的分类结果,必须充分利用高光谱图像的光谱信息和空间信息,减少类内的光谱变化,并扩大类别间的光谱差异。为此,提出一种滚动引导递归滤波的高光谱图像光谱—空间分类方法。首先,利用主成分分析对高光谱图像进行降维;然后,利用高斯滤波对输入图像进行模糊化,消除图像中的噪声和小尺度结构;接下来,将模糊化后的图像作为引导图像,对输入图像进行边缘保持递归滤波,输出结果作为新的引导图像,重复迭代这个过程直至大尺度边缘被恢复;最后,利用提取的特征波段和支持向量机对高光谱图像进行分类。在两个真实高光谱数据集上进行了分类实验,结果表明本文方法的分类精度优于其他的高光谱图像分类方法。在训练样本极少的情况下,本文方法也能获得较高的分类精度。  相似文献   

基于组分反演的高光谱潮间带表层沉积物自动分类研究   总被引：4，自引：2，他引：2  

张东  张鹰  李欢  陈晔 《海洋学报》2008,30(3):62-71

利用星载高光谱遥感影像数据Hyperion并结合现场采样粒度分析资料,对潮间带表层沉积物类型的自动分类进行了研究.根据低潮位下沉积物的短波红外光谱反射率与沉积物组分(沙、粉砂、黏土)含量之间的强相关关系,选择2102nm为特征波段,建立了高光谱定量遥感反演模型,获取初步的逐像元沙、粉砂和黏土含量,然后采用线性均衡化处理方法对初步反演结果进行修正,并结合Shepard三角分类法建立对照表,自动确定出每一像元对应的沉积物类型,得到研究区潮间带表层沉积物的空间分布特征,总体分类验证精度为87.9%.研究结果表明,研究区分布着三种沉积物,分别为粉砂质砂、砂质粉砂和粉砂,所占比例分别为18.9%,64.6%和16.5%,它们的空间分布趋势与研究区内的植被分布状况、水动力条件以及沉积物来源密切相关.  相似文献