Remote sensing image classification with GIS data based on spatial data mining techniques

Data mining techniques are used to discover knowledge from GIS database in order to improve remote sensing image classification. Two learning granularities are proposed for inductive learning from spatial data, one is spatial object granularity, the other is pixel granularity. We also present an approach to combine inductive learning with conventional image classification methods, which selects class probability of Bayes classification as learning attributes. A land use classification experiment is performed in the Beijing area using SPOT multi-spectral image and GIS data. Rules about spatial distribution patterns and shape features are discovered by C5.0 inductive learning algorithm and then the image is reclassified by deductive reasoning. Comparing with the results produced only by Bayes classification, the overall accuracy increased by 11% and the accuracy of some classes, such as garden and forest, increased by about 30%. The results indicate that inductive learning can resolve spectral confusion to a great extent. Combining Bayes method with inductive learning not only improves classification accuracy greatly, but also extends the classification by subdividing some classes with the discovered knowledge.     

基于空间数据发掘的遥感图像分类方法研究

采用数据发掘技术从 ＧＩＳ数据库和遥感图像中发现知识,用于改善遥感图像分类。提出了两种实施空间数据归纳学习的途径：在空间对象粒度上学习和直接在像元粒度上学习。分析了两种粒度学习的特点和适用范围,同时提出了一种归纳学习与传统图像分类法的结合方式。用北京地区 ＳＰＯＴ多光谱图像和 ＧＩＳ数据库进行土地利用分类的试验证明,归纳学习能较好地解决同谱异物、同物异谱等问题,显著提高分类精度,并且能够根据发现的知识进一步细分类,扩展了遥感图像分类的能力。     

GIS辅助下的Bayes法遥感影像分类

介绍了Bayes分类器 ,提出了从GIS空间数据库中挖掘知识用以辅助进行遥感影像分类的方法。文中以规则的形式表示遥感影像的解译知识 ,并使用其它地理辅助数据 ,从遥感影像处理、地理辅助数据、专家知识一体化的角度出发 ,使用基于知识的方法进行了分类研究 ,改善了分类精度。实验表明这是一种较好的分类方法。     

地理信息系统支持下的知识获取及其在遥感影像植被分类中的应用研究

本文根据植被类型分布与地理环境因子的关系,在地理信息系统和遥感技术支持下,通过ＧＩＳ叠加、统计分析操作,建立植被分布与年积温、降水量、海拔高度、土壤类型等环境因子的定量化知识向量表。综合应用所得到的地学知识向量表和植被光谱特征值进行分类试验,得到研究区的植被分布图。文章以贺兰山地区为例,详细介绍该方法的应用。     

特征提取策略对高分辨率遥感图像场景分类性能影响的评估

高分辨率遥感图像场景分类方法主要涉及两个环节:特征提取以及特征分类,分类器的设计已经相对成熟,当前工作的重点是特征提取策略的研究。为了进一步推动特征提取策略的研究,将特征提取策略对高分辨率遥感图像场景分类性能的影响进行了定性和定量评估。首先,回顾了高分辨率遥感图像场景分类的发展历程;然后,对现有高分辨率遥感图像场景分类方法的特征提取策略进行分类总结,并从理论上将各类特征提取策略对场景分类性能的影响进行定性评估;最后,在3个规模较大的数据集上对多种特征提取策略进行实验对比,将不同特征提取策略对场景分类性能的影响和各数据集的复杂度进行定量评估。     

A new Likelihood Ratio for supervised classification of fully polarimetric SAR data: An application for sea ice type mapping

One of the potential applications of polarimetric Synthetic Aperture Radar (SAR) data is the classification of land cover, such as forest canopies, vegetation, sea ice types, and urban areas. In contrast to single or dual polarized SAR systems, full polarimetric SAR systems provide more information about the physical and geometrical properties of the imaged area. This paper proposes a new Bayes risk function which can be minimized to obtain a Likelihood Ratio (LR) for the supervised classification of polarimetric SAR data. The derived Bayes risk function is based on the complex Wishart distribution. Furthermore, a new spatial criterion is incorporated with the LR classification process to produce more homogeneous classes. The application for Arctic sea ice mapping shows that the LR and the proposed spatial criterion are able to provide promising classification results. Comparison with classification results based on the Wishart classifier, the Wishart Likelihood Ratio Test Statistic (WLRTS) proposed by Conradsen et al. (2003) and the Expectation Maximization with Probabilistic Label Relaxation (EMPLR) algorithm are presented. High overall classification accuracy of selected study areas which reaches 97.8% using the LR is obtained. Combining the derived spatial criterion with the LR can improve the overall classification accuracy to reach 99.9%. In this study, fully polarimetric C-band RADARSAT-2 data collected over Franklin Bay, Canadian Arctic, is used.     

领域知识辅助下基于多尺度与主方向纹理的遥感影像土地利用分类

基于灰度共生矩阵(GLCM)的纹理特征在影像空间分析中具有重要作用,提出了一种在领域空间知识辅助下构建GLCM多尺度窗口与主方向权值的方法,从而提高纹理特征的有效性,并解决影像土地利用分类中存在的不确定性问题。为此,根据人类目视解译的特点,对GIS与RS数据进行集成计算:首先,在图像配准的基础上,利用经典的GIS空间数据挖掘算法,渐近式地提取领域形态知识;接着,采用关联分析法建立其与GLCM构造因子之间的响应机制,并设计了基于地类形状指数的多尺度窗口建立算法,以及基于地类主方向分布指数的方向权值测度算法。试验结果表明,领域形态知识与GLCM空间因子之间具有强相关关系,该方法提取出的纹理特征可以描述复杂地物的空间意义,算法复杂度低,性能优越,有效提高了影像土地利用分类的精度。     

基于分类规则挖掘的遥感影像分类研究

分析了目前遥感影像的统计分类、神经网络分类及基于符号知识的逻辑推理分类方法的优缺点.以GIS为平台,构建了多源空间数据库,将数据挖掘的思想和方法引入遥感影像分类中,提出了面向分类规则挖掘的遥感影像分类框架.针对遥感光谱数据及其他空间数据的特点,定义了连续属性样本分类概念和分割点评价指标,提出了一种新的连续属性样本分类规则挖掘算法.选择一个试验区,采用该算法分别对遥感光谱数据、遥感光谱和DEM数据相结合的数据进行分类规则挖掘、遥感影像分类和分类精度比较.结果表明：（1）该算法具有较高的分类精度;（2）加入DEM等与分类相关的其他空间数据可以提高遥感影像的分类精度.通过挖掘分类规则进行遥感影像分类,扩展了基于知识的逻辑推理分类方法中知识获取渠道,提高了分类规则获取的智能化程度.新的连续属性样本分类规则挖掘算法,扩展了归纳学习算法对连续属性样本分类的适应性.     

Study on gis data directly integrated into image processing

The basic properties to be dealt with ,when considering initially the research needs related to the integration of remote sensing (RS) information into a geographic information system (GIS),are many-sided.The primary combination of remote sensing and GIS is mainly realized by the transforms of data structure .Because of its own limitations,there is an urgent need to investigate the integration of RS and GIS in higher levels.In this paper,we discuss the different types of combinations of RS with GIS,and propose that GIS data should be directly brought into image processing from the beginning.A tentative idea of how to use the method of granularity to study the common processing unit of RS and GIS is described.The example for the determination of granularity of spatial data processing related to run-length-code line is also given.     

Study on gis data directly integrated into image processing

The basic properties to be dealt with, when considering initially the research needs related to the integration of remote sensing (RS) information into a geographic information system (GIS), are many-sided. The primary combination of remote sensing and GIS is mainly realized by the transforms of data structure. Because of its own limitations, there is an urgent need to investigate the integration of RS and GIS in higher levels. In this paper, we discuss the different types of combinations of RS with GIS, and propose that GIS data should be directly brought into image processing from the beginning. A tentative idea of how to use the method of granularity to study the common processing unit of RS and GIS is described. The example for the determination of granularity of spatial data processing related to run-length-code line is also given.     

Comparative performance of convolutional neural network,weighted and conventional support vector machine and random forest for classifying tree species using hyperspectral and photogrammetric data

ABSTRACT

The classification of tree species can significantly benefit from high spatial and spectral information acquired by unmanned aerial vehicles (UAVs) associated with advanced classification methods. This study investigated the following topics concerning the classification of 16 tree species in two subtropical forest fragments of Southern Brazil: i) the potential integration of UAV-borne hyperspectral images with 3D information derived from their photogrammetric point cloud (PPC); ii) the performance of two machine learning methods (support vector machine – SVM and random forest – RF) when employing different datasets at a pixel and individual tree crown (ITC) levels; iii) the potential of two methods for dealing with the imbalanced sample set problem: a new weighted SVM (wSVM) approach, which attributes different weights to each sample and class, and a deep learning classifier (convolutional neural network – CNN), associated with a previous step to balance the sample set; and finally, iv) the potential of this last classifier for tree species classification as compared to the above mentioned machine learning methods. Results showed that the inclusion of the PPC features to the hyperspectral data provided a great accuracy increase in tree species classification results when conventional machine learning methods were applied, between 13 and 17% depending on the classifier and the study area characteristics. When using the PPC features and the canopy height model (CHM), associated with the majority vote (MV) rule, the SVM, wSVM and RF classifiers reached accuracies similar to the CNN, which outperformed these classifiers for both areas when considering the pixel-based classifications (overall accuracy of 84.4% in Area 1, and 74.95% in Area 2). The CNN was between 22% and 26% more accurate than the SVM and RF when only the hyperspectral bands were employed. The wSVM provided a slight increase in accuracy not only for some lesser represented classes, but also some major classes in Area 2. While conventional machine learning methods are faster, they demonstrated to be less stable to changes in datasets, depending on prior segmentation and hand-engineered features to reach similar accuracies to those attained by the CNN. To date, CNNs have been barely explored for the classification of tree species, and CNN-based classifications in the literature have not dealt with hyperspectral data specifically focusing on tropical environments. This paper thus presents innovative strategies for classifying tree species in subtropical forest areas at a refined legend level, integrating UAV-borne 2D hyperspectral and 3D photogrammetric data and relying on both deep and conventional machine learning approaches.     

基于随机森林算法的多维情境特征活动识别

利用智能手机传感器可感知时间、空间、时空和用户等多维情境的特征,可识别用户活动,但原框架模型中仅利用了单一分类器中的朴素贝叶斯算法,存在分类精度效果受限的问题。本文利用集成分类器中的随机森林算法对原有框架中的单一分类器进行了改进。在获取的3个数据集上的十倍交叉验证结果表明,加权平均F1量测值均有较大提高,表明利用随机森林算法在分类精度效果上有所提升;但由于集成算法结构相对复杂,其学习效率相对较低。此外,随机森林算法的分类混淆矩阵表明,导致识别误差的因素主要为活动的定义与室内定位精度。     

基于Sentinel-1A数据的多种机器学习算法识别冰山的比较

冰山识别对于海洋环境监测和船只安全运行等具有重要的意义,是北极航道开通和北极开发过程中的重要内容。采用合成孔径雷达(SAR)影像进行冰山识别具有独特的优势,多种机器学习算法均可用于SAR影像的冰山识别中。为了最大限度地发挥机器学习算法的性能,有必要对不同机器学习算法及其搭配使用的特征与特征标准化方法进行评估,从而进行最优冰山识别方法的选择。因此,本文基于Sentinel-1A SAR影像,采用多种机器学习方法、多种特征组合及多种特征标准化方法进行冰山识别,并比较各流程方法的识别性能差异。采用的机器学习算法包括贝叶斯分类器(Bayes)、反向神经网络(BPNN)、线性判别分析(LDA)、随机森林(RF)以及支持向量机(SVM);特征标准化方法包括Min-max标准化、Z-score标准化及log函数标准化;数据集是含有12个SAR影像特征的969个冰山与非冰山样本,样本主要位于格陵兰岛东海岸。分类效果采用接收者操作特性(ROC)曲线下的面积(AUC)进行衡量。结果显示,最佳搭配下的RF的AUC值最高,达到了0.945,比最差的Bayes高出0.09。从识别率上来看,RF在冰山查全率为80%的情况下非冰山查全率达到92.6%,效果最好,比第2位的BPNN高出1.4%,比最差的Bayes高出2.6%;BPNN在冰山查全率为90%的情况下非冰山查全率达到87.4%,比第2位的RF高出0.8%,比最差的Bayes高出2.7%。上述结果表明,对冰山识别而言,选择最优的机器学习算法和最佳的特征与特征标准化方法都是十分重要的。     

遥感影像分类的归类学习方法

遥感影像分类专家系统是遥感分类研究中的一个重要发展方向,然而,传统的统计模式识别法和人工神经网络分类法除了能完成具体的影像分类外,不能提供易于被人类理解的分类知识,文中介绍一种基于扩张矩阵的示例学习方法,并将其应用于遥感影像分类知识的自动获取。     

联合卷积神经网络与集成学习的遥感影像场景分类

针对人工设计的中、低层特征难以实现复杂场景影像的高精度分类以及卷积神经网络依赖大量训练数据等问题，结合迁移学习与集成学习，提出了一种联合卷积神经网络与集成学习的遥感影像场景分类算法。首先基于迁移学习的思想，利用在自然影像数据集上训练好的多个深层卷积神经网络模型作为特征提取器，提取图像多个高度抽象的语义特征；然后构建由Logistic回归和支持向量机组成的Stacking集成模型，对同一图像的多个特征分别训练Logistic模型，将预测概率结果融合构建概率特征；最后利用支持向量机对概率特征训练和预测，得到场景影像的分类结果。利用UCMerced_LandUse和NWPU-RESISC 45两种不同规模的遥感影像数据集进行试验，即使在只有10%的数据作为训练样本情况下，本文方法能够分别达到90.74%和87.21%的分类精度。     

Multiple band digital orthophoto quadrangle data: A source for generating land use maps

Abstract

The purpose of this study was to investigate the use of color infrared‐digital orthophoto quadrangle (CIR‐DOQ) data to generate land use/land cover (LULC) maps and to incorporate them as data layers in geographic information systems (GIS) involving various resource management scenarios. The Danville 7.5‐minute quadrangle located in the southern part of Limestone and Morgan counties, Alabama, was used as the study site. Data for the special CIR‐DOQ were generated by scanning four 9x9 inch CIR aerial photographs at a uniform pixel sample grid of 25 microns resulting in 2 meters ground sample resolution. One‐half of the quadrangle was used to identify training sites for performing a supervised classification of the data and the other half to verify the accuracy of the classification. The CIR‐DOQ data were found to be adequate for using a supervised classification algorithm to differentiate major LULC classes, resulting in a classification accuracy of 93 percent. The superior spatial quality of the data over commençai satellite data affords resource managers an opportunity to more effectively study land cover and surface hydrological properties of an area, soil moisture and surface soil textures, as well as differentiate among vegetation species, using remote sensing techniques. However, caution must be exercised when using multispectral classification techniques to classify mosaicked CIRDOQ data because of the image enhancements used to generate the final product. In its present form, there are some limitations to the use of the data for performing spectral classifications. Hozvever, the high spatial resolution of the data enables even the novice resource planner to effectively use the data in visual interpretations of major LULC classes.     

基于超像素图卷积网络的高光谱图像分类

高光谱图像分类是遥感领域中一个具有挑战性的问题。基于深度学习框架的高光谱图像分类方法,由于其良好的分类性能受到了越来越多的关注。然而,这些方法普遍存在的问题为：模型的训练不仅需要大量的时间,而且还需要大量的标签样本。针对此问题,本文提出了一种基于超像素图卷积网络的高光谱图像分类方法。该方法以超像素作为图的节点,极大地减小了图的规模,从而提高了分类效率;提出的超像素合并技术能有效地融合光谱-空间信息,增强了空间信息在分类中的作用;为了验证该方法的有效性,在Indian Pines、Pavia University两个实际数据集上进行试验,并与一些先进的基于深度学习框架的高光谱图像分类方法进行比较。结果表明,本文方法在分类精度和分类效率上均优于其他方法。     

Applications of remote sensing and geographic information systems for urban land-cover change studies in Mongolia

The aim of this study is to compare the changes that occurred in the main urban land-cover classes of Ulaanbaatar city, Mongolia, during a centralized economy with those that occurred during a market economy and to describe the socio-economic reasons for the changes. For this purpose, multi-temporal remote sensing and geographical information system (GIS) data sets, as well as census data, are used. To extract the reliable urban land-cover information from the selected remotely sensed data sets, a refined parametric classification algorithm that uses spatial thresholds defined from local and contextual knowledge is constructed. Before applying the classification decision rule, some image fusion techniques are applied to the selected remotely sensed data sets to define the most efficient fusion method for training sample selection and for defining local and contextual knowledge. Overall, the study indicates that during the centralized economy significant changes occurred in a ger area of the city, whereas during the market economy the changes occurred in all areas.     

Optimisation of building detection in satellite images by combining multispectral classification and texture filtering

Conventional multispectral classification methods show poor performance with respect to detection of urban object classes, such as buildings, in high spatial resolution satellite images. This is because objects in urban areas are very complicated with respect to both their spectral and spatial characteristics. Multispectral classification detects object classes only according to the spectral information of the individual pixels, while a large amount of spatial information is neglected. In this study, a technique is described which attempts to detect urban buildings in two stages. The first stage is a conventional multispectral classification. In the second stage, the classification of buildings is improved by means of their spatial information through a modified co-occurrence matrix based filtering. The direction dependence of the co-occurrence matrix is utilised in the filtering process. The method has been tested by using TM and SPOT Pan merged data for the whole area of the city of Shanghai, China. After the co-occurrence matrix based filtering, the average user accuracy increased by about 46% and the average Kappa statistic by about 57%. This result is about 26% better than the accuracy improvement through normal texture filtering. The method presented in this study is very useful for a rapid estimation of urban building and city development, especially in metropolitan areas of developing countries.     

用ESDA技术从GIS数据库中发现知识

从GIS数据库中可发现许多知识，这在GIS界已引起了相当的重视。ESDSA技术注重研究数据的空间依赖与空间异质性，在知识发现中用于选取感兴趣的数据子集，并可初步发现隐含在数据中的某些特征和规律。一些标准全程和局部空间统计包括MoranI,GearyG,G统计以及LISA等是ESDA技术的基本核心内容之一。在目前专家系统尚不成熟的条件下，充分利用GIS的可视化和空间分析技术是实现ESDA技术与GIS紧密完全结合的关系。实例说明在ESDA技术结合其它相关领域的知识从GIS数据库中发现知识的方法是可行的。