面向对象规则和支持向量机的天宫一号高光谱影像分类

传统的高光谱分类方法通常基于单一像元的光谱或纹理特征,很少考虑地物空间结构信息与空间相关特征.本文将面向对象规则与基于像元的分类进行融合,利用对象的空间结构特征和光谱特征进行混合分类,旨在克服像元层次分类的不足.本文尝试性的提出了两种混合分类方法:(1)基于分形网络演化的多尺度分割支持向量机分类(2)基于多层分水岭分割的SVM分类,并将这两种方法应用到天宫一号高光谱数据上.结果表明:基于面向对象规则的混合分类方法有效地提高了分类精度,不仅能够改善同谱异物现象,而且解决分类结果中地物破碎的问题.     

引入商空间粒度计算的全极化SAR影像分类

为了充分利用不同极化特征信息,并将其有效地结合,提出一种结合粒度计算的全极化合成孔径雷达（synthetic aperture radar,SAR）影像分类方法。在不同极化目标分解特征组合的基础上引入影像纹理信息,利用光滑支持向量机（smooth support vector machine,SSVM）对不同特征组合进行类别划分获得粗粒度空间,采用商空间对粗粒度进行合并;根据全极化SAR影像分布特性,以相干矩阵作为新的特征矢量,利用Wishart测度代替传统欧氏距离对差异粒度进行推理,通过合并推理结果与合成论域,获得精细分类结果。采用L波段San Francisco地区和荷兰Flevoland地区的全极化SAR影像进行分类试验,结果表明:利用SSVM算法对全极化SAR影像进行粗粒度划分,并采用Wishart距离对差异粒度推理综合,总体分类效果优于结合纹理信息的Cloude及Yamaguchi4分类结果,且优于基于线性特征融合进行监督分类方法。     

不同降维策略下的高光谱影像多特征分类

由于物体表面的空间分布通常是富有规律且局部连续的,在高光谱影像分类中应充分利用其光谱和空间信息。本文在对高光谱影像立方体进行降维处理的基础上,提出了一种联合空域和谱域信息的高光谱影像高效分类方法。首先,分别选用主成分分析(Principal Component Analysis,PCA)和正交投影波段选择(Orthogonal Projection Band Selection,OPBS)两种方法对原始高光谱数据进行预处理,获取降维后的影像数据。然后在其基础上提取扩展形态学特征(Extended Morphology Profiles,EMP)和地物表面纹理特征,组成联合光谱和纹理、形状结构特征。最后,采用支持向量机(Support Vector Machine,SVM)分类器对联合特征进行分类。针对不同真实高光谱数据集的实验结果表明,本文提出的方法运算效率高且具有令人满意的分类性能。     

基于支持向量机的CBERS-02卫星影像信息提取

CBERS卫星是由中国空间技术研究院与巴西空间研究院联合研制的地球资源遥感卫星,CBERS-02卫星数据总体质量比CBERS-01卫星有所提高,本文利用支持向量机方法对CBERS-02卫星影像信息进行提取。研究中首先用6S模式对影像进行大气校正,然后选择RBF为支持向量机方法的核函数,并用交叉验证方法得到影响RBF核函数的两个最佳参数值进行学习完成信息提取,最后将提取结果制作成矢量图。通过研究得出用大气校正后的数据进行信息提取分类精度有所提高;与最大似然法和最小距离法相比,支持向量机方法分类精度较高。通过将研究结果与ETM+影像进行比较得出,CBERS-02卫星影像精度能够满足应用需求并能代替TM/ETM+数据开展研究工作。     

支持向量机在空间信息处理领域的应用研究

支持向量机作为一种最新的也是最有效的统计学习方法,近年来成为模式识别与机器学习领域一个新的研究热点。支持向量机具有小样本学习、抗噪声性能好、学习效率高和推广性好的优点,能够用于空间信息处理分析领域的遥感影像处理、高光谱分类、拟合与回归、数据挖掘、目标检测等任务。本文在总结分析近年来支持向量机在空间信息处理领域应用主要进展与成果的基础上,结合支持向量机理论方法与空间信息处理的发展趋势,提出了今后有必要重点研究的若干问题,包括空间数据挖掘、智能空间信息处理、高维空间数据处理等。     

GPS高程拟合支持向量机模型

为了快速获取GPS高程异常值,提出了基于最小二乘支持向量机（LS—SVM）的GPS高程异常值求取模型,介绍了最小二乘支持向量机的原理与优越性。利用该模型进行了高程异常的拟合,并对已知点进行了检验。结果表明：其结果是可靠的,在有限样本情况下完全可以达到传统GPS高程拟舍的效果,且其实现起来更简单,具有一定的科学性和实用性。     

Vegetation species mapping in a coastal-dune ecosystem using high resolution satellite imagery

Vegetation mapping is a priority when managing natural protected areas. In this context, very high resolution satellite remote sensing data can be fundamental in providing accurate vegetation cartography at species level. In this work, a complete processing methodology has been developed and validated in a complex vulnerable coastal-dune ecosystem. Specifically, the analysis has been carried out using WorldView-2 imagery, which offers spatial and spectral resolutions. A thorough assessment of 5 atmospheric correction models has been performed using real reflectance measures from a field radiometry campaign. To select the classification methodology, different strategies have been evaluated, including additional spectral (23 vegetation indices) and spatial (4 texture parameters) information to the multispectral bands. Likewise, the application of linear unmixing techniques has been tested and abundance maps of each plant species have been generated using the library of spectral signatures recorded during the campaign. After the analysis conducted, a new methodology has been proposed based on the use of the 6S atmospheric model and the Support Vector Machine classification algorithm applied to a combination of different spectral and spatial input data. Specifically, an overall accuracy of 88,03% was achieved combining the corrected multispectral bands plus a vegetation index (MSAVI2) and texture information (variance of the first principal component). Furthermore, the methodology has been validated by photointerpretation and 3 plant species achieve significant accuracy: Tamarix canariensis (94,9%), Juncus acutus (85,7%) and Launaea arborescens (62,4%). Finally, the classified procedure comparing maps for different seasons has also shown robustness to changes in the phenological state of the vegetation.     

基于波段子集的独立分量分析的特征提取的高光谱遥感影像分类

针对高光谱影像数据具有波段众多、数据量较大的特点,本文提出了一种基于波段子集的独立分量分析(ICA)特征提取的高光谱遥感影像分类的新方法。以北京昌平小汤山地区的高光谱影像为例,根据高光谱遥感影像的相邻波段的相关性进行子空间划分,在各个波段子集上采用ICA算法进行特征提取,将各个子空间提取的特征合并组成特征向量,采用支持向量机(SVM)分类器进行分类。结果表明:该方法分类精度最佳(分类精度89.04%,Kappa系数0.8605,明显优于其它特征提取方法的SVM分类,有效地提高了高光谱数据的分类精度。     

Novel Folded-PCA for improved feature extraction and data reduction with hyperspectral imaging and SAR in remote sensing

As a widely used approach for feature extraction and data reduction, Principal Components Analysis (PCA) suffers from high computational cost, large memory requirement and low efficacy in dealing with large dimensional datasets such as Hyperspectral Imaging (HSI). Consequently, a novel Folded-PCA is proposed, where the spectral vector is folded into a matrix to allow the covariance matrix to be determined more efficiently. With this matrix-based representation, both global and local structures are extracted to provide additional information for data classification. Moreover, both the computational cost and the memory requirement have been significantly reduced. Using Support Vector Machine (SVM) for classification on two well-known HSI datasets and one Synthetic Aperture Radar (SAR) dataset in remote sensing, quantitative results are generated for objective evaluations. Comprehensive results have indicated that the proposed Folded-PCA approach not only outperforms the conventional PCA but also the baseline approach where the whole feature sets are used.     

The effect of atmospheric and topographic correction on pixel-based image composites: Improved forest cover detection in mountain environments

Quantification of forest cover is essential as a tool to stimulate forest management and conservation. Image compositing techniques that sample the most suited pixel from multi-temporal image acquisitions, provide an important tool for forest cover detection as they provide alternatives for missing data due to cloud cover and data discontinuities. At present, however, it is not clear to which extent forest cover detection based on compositing can be improved if the source imagery is firstly corrected for topographic distortions on a pixel-basis. In this study, the results of a pixel compositing algorithm with and without preprocessing topographic correction are compared for a study area covering 9 Landsat footprints in the Romanian Carpathians based on two different classifiers: Maximum Likelihood (ML) and Support Vector Machine (SVM). Results show that classifier selection has a stronger impact on the classification accuracy than topographic correction. Finally, application of the optimal method (SVM classifier with topographic correction) on the Romanian Carpathian Ecoregion between 1985, 1995 and 2010 shows a steady greening due to more afforestation than deforestation.     

似大地水准面的支持向量机模型研究

支持向量机(SVM)是一种基于结构风险最小化原理的学习技术,也是一种新的具有较好泛化性能的回归方法。本文简要介绍了SVM原理,针对大面积复杂似大地水准面的确定问题,仅依据测区的GPS水准实测数据,利用SVM方法整体建模。通过工程实例并与神经网络模型进行对比,证实了SVM似大地水准面模型的可靠性。     

基于SVM的遥感影像的分类

传统的遥感图像的分类方法包括统计模式识别、句法模式识别、以及神经网络、遗传算法、模拟退火算法等。分析了统计模式识别的方法的优缺点,提出了使用SVM的方法进行遥感图像分类的设想,通过实验证明该方法是有效的和稳健的。     

Corn Area Extraction by the Integration of MODIS-EVI Time Series Data and China’s Environment Satellite (HJ-1) Data

In this paper, we intent to use the remotely sensed MODerate resolution Imaging Spectroradiometer (MODIS) data and China’s Environment Satellite (HJ-1) data for extracting the corn cultivated area over a regional scale. The high resolution HJ-1 data was to extract corn distribution at a small scale class with Support Vector Machine (SVM). The mean Enhanced Vegetation Index (EVI) time series curve of corn from MODIS was derived for the reference area and validated in a larger area. The MODIS-EVI time series curve derived from the reference area instead of the MODIS-EVI time series curve derived from the study area after validation, which was taken as the standard MODIS-EVI time series curve in for generating a standard MODIS-EVI image of corn. The mean absolute distance (MAD) between the standard MODIS-EVI image of corn and the MODIS-EVI time series image was used to detect the maximum possible extent of corn distribution in the study area. The results showed that the overall accuracy of the method was 82.17 %, with commission and omission errors of 16.85 and 15.40 %, respectively; at the county level, the satellite-estimated corn area and statistical data were well correlated (R ²?=?0.85, N?=?50) for the whole Jilin Province. It indicated that the MODIS data integrated with higher spatial resolution of HJ-1 satellite data could be utilized to enhance the extraction accuracy of corn cultivated area at a larger scale.     

A Multiple SVM System for Classification of Hyperspectral Remote Sensing Data

With recent technological advances in remote sensing sensors and systems, very high-dimensional hyperspectral data are available for a better discrimination among different complex land-cover classes. However, the large number of spectral bands, but limited availability of training samples creates the problem of Hughes phenomenon or ‘curse of dimensionality’ in hyperspectral data sets. Moreover, these high numbers of bands are usually highly correlated. Because of these complexities of hyperspectral data, traditional classification strategies have often limited performance in classification of hyperspectral imagery. Referring to the limitation of single classifier in these situations, Multiple Classifier Systems (MCS) may have better performance than single classifier. This paper presents a new method for classification of hyperspectral data based on a band clustering strategy through a multiple Support Vector Machine system. The proposed method uses the band grouping process based on a modified mutual information strategy to split data into few band groups. After the band grouping step, the proposed algorithm aims at benefiting from the capabilities of SVM as classification method. So, the proposed approach applies SVM on each band group that is produced in a previous step. Finally, Naive Bayes (NB) as a classifier fusion method combines decisions of SVM classifiers. Experimental results on two common hyperspectral data sets show that the proposed method improves the classification accuracy in comparison with the standard SVM on entire bands of data and feature selection methods.     

大坝变形预测的支持向量机模型

针对大坝变形具有强非线性的特点以及在采用传统神经网络模型进行预测时存在局部极小、过学习等问题,提出一种新的大坝变形预测方法——支持向量机方法。该方法基于统计学习理论,采用结构风险最小化原则,保证了模型具有很强的泛化性能,并通过求解一个二次规划问题确保模型具有全局最优。以东江大坝变形预测为实例,说明了该方法的可行性和有效性。     

Support Vector Machine for Spatial Variation

Large, multivariate geographic datasets have been used to characterize geographic space with the help of spatial data mining tools. In our study, we explore the sufficiency of the Support Vector Machine (SVM), a popular machine‐learning technique for unsupervised classification and clustering, to help recognize hidden patterns in a college admissions dataset. Our college admissions dataset holds over 10,000 students applying to an undisclosed university during one undisclosed year. Students are qualified almost exclusively by their standardized test scores and school records, and a known admissions decision is rendered based on these criteria. Given that the university has a number of political, social and geographic econometric factors in its admissions decisions, we use SVM to find implicit spatial patterns that may favor students from certain geographic regions. We first explore the characteristics of the applicants in the college admissions case study. Next, we explain the SVM technique and our unique ‘threshold line’ methodology for both discrete (regional) and continuous (k‐neighbors) space. We then analyze the results of the regional and k‐neighbor tests in order to respond to the methodological and geographic research questions.     

Development of fuzzy rule-based parameters for urban object-oriented classification using very high resolution imagery

Urban areas consist of spectrally and spatially heterogeneous features. Advanced information extraction techniques are needed to handle high resolution imageries in providing detailed information for urban planning applications. This study was conducted to identify a technique that accurately maps impervious and pervious surfaces from WorldView-2 (WV-2) imagery. Supervised per-pixel classification algorithms including Maximum Likelihood and Support Vector Machine (SVM) were utilized to evaluate the capability of spectral-based classifiers to classify urban features. Object-oriented classification was performed using supervised SVM and fuzzy rule-based approach to add spatial and texture attributes to spectral information. Supervised object-oriented SVM achieved 82.80% overall accuracy which was the better accuracy compared to supervised per-pixel classifiers. Classification based on the proposed fuzzy rule-based system revealed satisfactory output compared to other classification techniques with an overall accuracy of 87.10% for pervious surfaces and an overall accuracy of 85.19% for impervious surfaces.     

天宫一号高光谱数据的船只检测实验

针对检测目标进行基于引力场的船只目标增强,计算船只目标的纹理、光谱等指标,基于支持向量机(SVM)方法,利用天宫一号高光谱数据,得到空间探测信息与地面实际目标之间存在精确的相互关系,进行船只检测实验研究.结果表明:天宫一号高光谱数据能够有效的从海水背景中提取船只信息,但船只类型仍难以有效识别.图像的自动分割对于船只的检测非常重要.传感器的多光谱与高分辨率相结合是今后研制的一个主要方向.     

Grassland habitat mapping by intra-annual time series analysis – Comparison of RapidEye and TerraSAR-X satellite data

Remote sensing concepts are needed to monitor open landscape habitats for environmental change and biodiversity loss. However, existing operational approaches are limited to the monitoring of European dry heaths only. They need to be extended to further habitats. Thus far, reported studies lack the exploitation of intra-annual time series of high spatial resolution data to take advantage of the vegetations’ phenological differences. In this study, we investigated the usefulness of such data to classify grassland habitats in a nature reserve area in northeastern Germany. Intra-annual time series of 21 observations were used, acquired by a multi-spectral (RapidEye) and a synthetic aperture radar (TerraSAR-X) satellite system, to differentiate seven grassland classes using a Support Vector Machine classifier. The classification accuracy was evaluated and compared with respect to the sensor type – multi-spectral or radar – and the number of acquisitions needed. Our results showed that very dense time series allowed for very high accuracy classifications (>90%) of small scale vegetation types. The classification for TerraSAR-X obtained similar accuracy as compared to RapidEye although distinctly more acquisitions were needed. This study introduces a new approach to enable the monitoring of small-scale grassland habitats and gives an estimate of the amount of data required for operational surveys.     

半监督条件随机场的高光谱遥感图像分类

善于捕捉空间信息的条件随机场模型虽然已被应用于高光谱遥感图像分类,但条件随机场的性能受到了标注训练样本数量的制约。为解决上述问题,本文提出了一种半监督条件随机场模型用于高光谱遥感图像分类。在该模型中,首先,利用空间-光谱拉普拉斯支持向量机定义关联势函数,以利用未标注样本中包含的信息获取样本类别概率;然后,在交互势函数中嵌入未标注的空间邻域样本,以充分利用空间信息实现对样本类别概率的修正;最后,采用分布式学习策略和平均场完成半监督条件随机场的训练和推断。本文在两个公开的高光谱数据集(Indian Pines数据集,Pavia University数据集)上进行了实验。实验结果表明Kappa系数提升3.94%。