结合多分类器的遥感数据专题分类方法研究

采用标准的多分类器结合方法进行遥感图像的分类研究。首先介绍了标准的多分类器结合的算法,然后以Landsat-TM多光谱遥感数据的土地覆被分类为例,分别给出了抽象级上相同训练特征的多分类器结合、抽象级上不同训练特征的多分类器结合和测量级上的多分类器结合进行土地覆被分类的方法,并进行了实例研究。参与分类器结合的单个分类器包括最大似然分类器,最小距离分类器,马氏距离分类器,K-NN分类器,多层感知器神经网络分类器。分类器的分类精度用总体精度、用户精度、生产者精度、kappa系数和条件kappa系数评价。结果表明,每一种多分类器结合的分类方法都能够比较显著地提高总体分类精度。文章最后对不同多分类器结合方式的优缺点进行了分析。     

Dual-tree complex wavelet transform-based image enhancement for accurate long-term change assessment in coal mining areas

The main objective of this study was to improve the long-term land use change detection by improving classification accuracy of previous generation satellite image using a recent super-resolution technique. The study also analysed the change in land cover over a period of 41 years in a coal mining area. A dual-tree complex wavelet transform-based image super-resolution technique was used to enhance Landsat images of 1975 and 2016. Separating pixels with similar spectral response is an enigmatical task, especially when those pixel represent different ground features. Therefore, an advanced neural net supervised classifier was used to minimize classification errors. Accuracy of the classified images (both super-resolved and original) were measured using confusion matrices and kappa coefficients. A significant improvement of more than 10% was observed in the overall classification accuracy for the image of 1975, highlighting that the classification accuracy of earlier generation satellite data can be improved substantially.     

Evaluation of different machine learning methods for land cover mapping of a Mediterranean area using multi-seasonal Landsat images and Digital Terrain Models

Land cover monitoring using digital Earth data requires robust classification methods that allow the accurate mapping of complex land cover categories. This paper discusses the crucial issues related to the application of different up-to-date machine learning classifiers: classification trees (CT), artificial neural networks (ANN), support vector machines (SVM) and random forest (RF). The analysis of the statistical significance of the differences between the performance of these algorithms, as well as sensitivity to data set size reduction and noise were also analysed. Landsat-5 Thematic Mapper data captured in European spring and summer were used with auxiliary variables derived from a digital terrain model to classify 14 different land cover categories in south Spain. Overall, statistically similar accuracies of over 91% were obtained for ANN, SVM and RF. However, the findings of this study show differences in the accuracy of the classifiers, being RF the most accurate classifier with a very simple parameterization. SVM, followed by RF, was the most robust classifier to noise and data reduction. Significant differences in their performances were only reached for thresholds of noise and data reduction greater than 20% (noise, SVM) and 25% (noise, RF), and 80% (reduction, SVM) and 50% (reduction, RF), respectively.     

Dimensionality Reduction and Classification of Hyperspectral Images Using Object-Based Image Analysis

Object-based image analysis (OBIA) has attained great importance for the delineation of landscape features, particularly with the accessibility to satellite images with high spatial resolution acquired by recent sensors. Statistical parametric classifiers have become ineffective mainly due to their assumption of normal distribution, vast increase in the dimensions of the data and availability of limited ground sample data. Despite pixel-based approaches, OBIA takes semantic information of extracted image objects into consideration, and thus provides more comprehensive image analysis. In this study, Indian Pines hyperspectral data set, which was recorded by the AVIRIS hyperspectral sensor, was used to analyse the effects of high dimensional data with limited ground reference data. To avoid the dimensionality curse, principal component analysis (PCA) and feature selection based on Jeffries–Matusita (JM) distance were utilized. First 19 principal components representing 98.5% of the image were selected using the PCA technique whilst 30 spectral bands of the image were determined using JM distance. Nearest neighbour (NN) and random forest (RF) classifiers were employed to test the performances of pixel- and object-based classification using conventional accuracy metrics. It was found that object-based approach outperformed the traditional pixel-based approach for all cases (up to 18% improvement). Also, the RF classifier produced significantly more accurate results (up to 10%) than the NN classifier.     

典型遥感数据分类方法的适用性分析——以遥感图像场景分类为例

目前普遍采用的分类器通常都是针对单一或小量任务而设计的,在小数据量的处理中能取得比较满意的结果。但对于海量遥感数据的处理,其在处理时效和分类精度方面还有待研究。本文以遥感图像场景分类任务为例,着重对遥感数据分类问题中几种典型分类方法的适用性进行比较研究,包括K近邻(KNN)、随机森林(RF),支持向量机(SVM)和稀疏表达分类器(SRC)等。分别从参数敏感性,训练样本数据量,待分类样本数据量和样本特征维数对分类器性能的影响等几个方面进行比较分析。实验结果表明:(1)KNN,RF和L0-SRC方法相比RBF-SVM,Linear-SVM和L1-SRC,受参数影响的程度更弱;(2)待分类样本固定的情况下,随着训练样本数目的增加,SRC类型分类方法的分类性能最佳,SVM类型方法次之,然后是RF和KNN,在总体分类时间上呈现出L0-SRCL1-SRCRFRBF-SVM/Linear-SVMKNN/L0-SRC-Batch的趋势;(3)训练样本固定的情况下,所有分类方法的分类精度几乎都不受待分类样本数目变化的影响,RBF-SVM方法性能最佳,其次是L1-SRC,然后是Linear-SVM,最后是RF和L0-SRC/L0-SRC-Batch,在总体分类时间上,L1-SRC和L0-SRC相比其他分类方法最为耗时;(4)样本特征维数的变化不仅影响分类器的运行效率,同时也影响其分类精度,其中SRC和KNN分类器器无需较高的特征维数即可获得较好的分类结果,SVM对高维特征具有较强的包容性和学习能力,RF分类器对特征维数增加则表现得并不敏感,特征维数的增加并不能对其分类精度的提升带来更多的贡献。总的来说,在大数据量的遥感数据分类任务中,现有分类方法具有良好的适用性,但是对于分类器的选择应当基于各自的特点和优势,结合实际应用的特点进行权衡和选择,选择参数敏感性较小,分类总体时间消耗低但分类精度相对较高的分类方法。     

顾及分类器参数的全极化SAR图像特征选择与分类

全极化SAR获取的信息量远多于传统SAR,但信息量的增加并不能确保分类精度的提高,如何有效进行特征选择至关重要。针对自适应特征选择问题,提出一种顾及分类器参数的特征选择和分类方法。该方法以支持向量数为评估依据,结合遗传算法进行特征选择,并同时对分类器参数进行寻优;最后利用优选的特征集和模型参数进行分类。为验证算法的有效性,利用两组全极化数据进行了监督分类实验。实验结果表明,提出方法降低了SVM分类器对自身参数的敏感性,而且能在较少特征个数下具备良好的泛化性能,分类精度优于未经过特征选择和参数优化的方法。     

遥感图像土地覆盖分类中多源特征数据选择研究

多源特征数据可以提高遥感图像的分类精度,选择合适的特征数据十分重要。利用基尼指数对多尺度纹理信息、主成分变换前三分量、地形数据等特征进行选择,选出最佳特征子集。利用支持向量机、神经网络分类法、最大似然法分别对全部特征数据和最佳特征子集结合多光谱数据进行分类。实验结果表明:基尼指数可以有效地对多源特征数据进行选择,特征选择可以提高分类器效率,提高分类精度。     

基于随机森林特征优选的冬小麦分类方法

本文基于多时相Landsat 8 OLI数据,进行综合光谱、植被指数的特征提取与特征选择的方法研究。通过分析光谱与植被指数特征时序变化,提取最佳时相光谱,构建小麦提取特征;采用基于重要性与Pearson相关性的随机森林特征选择算法优选特征。结果表明:利用优选特征分类时,总体精度为89.78%,小麦分类精度为98.33%;与优选前特征的分类结果相比,精度分别提高了2.96%、2.55%;基于重要性与Pearson相关性的随机森林特征选择提高了分类精度和分类器工作效率。     

A semi-ellipsoid-model based fuzzy classifier to map grassland in Inner Mongolia,China

Remote sensing techniques offer effective means for mapping plant communities. However, mapping grassland with fine vegetative classes over large areas has been challenging for either the coarse resolutions of remotely sensed images or the high costs of acquiring images with high-resolutions. An improved hybrid-fuzzy-classifier (HFC) derived from a semi-ellipsoid-model (SEM) is developed in this paper to achieve higher accuracy for classifying grasslands with Landsat images. The Xilin River Basin, Inner Mongolia, China, is chosen as the study area, because an acceptable volume of ground truthing data was previously collected by multiple research communities. The accuracy assessment is based on the comparison of the classification outcomes from four types of image sets: (1) Landsat ETM+ August 14, 2004, (2) Landsat TM August 12, 2009, (3) the fused images of ETM+ with CBERS, and (4) TM with CBERS, respectively, and by three classifiers, the proposed HFC-SEM, the tetragonal pyramid model (TPM) based HFC, and the support vector machine method. In all twelve classification experiments, the HFC-SEM classifier had the best overall accuracy statistics. This finding indicates that the medium resolution Landsat images can be used to map grassland vegetation with good vegetative detail when the proper classifier is applied.     

高光谱影像的引导滤波多尺度特征提取

为了解决高光谱遥感影像分类中单一尺度特征无法有效表达地物类间差异和区分地物边界的不足,提高影像分类精度和改善分类目视解译效果,提出了采用引导滤波提取多尺度的空间特征的方法。首先,利用主成分分析对高光谱影像进行降维,移除噪声并突出主要特征;然后,将第1主成分作为引导影像,将包含信息量最多的若干主成分分别作为输入影像,应用依次增加的滤波半径分别进行引导滤波处理提取多个尺度的特征,获得影像不同尺度的结构信息;最后,将多尺度特征输入分类器中进行影像监督分类。采用仿真数据和帕维亚大学(Pavia University)、帕维亚城区(Pavia Centre)等3幅高光谱实验数据,提取了基于引导滤波的多尺度特征、多尺度形态特征和多尺度纹理特征,输入到支持向量机、随机森林和K近邻分类器中,进行了实验。实验结果表明:采用支持向量机分类Pavia University数据,相对于采用多尺度形态特征的分类结果,引导滤波特征的总体精度提高了6.5%;Pavia Centre和Salinas两幅影像最高分类精度均由引导滤波特征实现,分别达到98.51%和98.39%。实验证实基于引导滤波提取的多尺度特征能有效地描述地物结构,进而获得更高的分类精度和改善目视解译效果。     

Object-based crop classification using multi-temporal SPOT-5 imagery and textural features with a Random Forest classifier

In this study, an object-based image analysis (OBIA) approach was developed to classify field crops using multi-temporal SPOT-5 images with a random forest (RF) classifier. A wide range of features, including the spectral reflectance, vegetation indices (VIs), textural features based on the grey-level co-occurrence matrix (GLCM) and textural features based on geostatistical semivariogram (GST) were extracted for classification, and their performance was evaluated with the RF variable importance measures. Results showed that the best segmentation quality was achieved using the SPOT image acquired in September, with a scale parameter of 40. The spectral reflectance and the GST had a stronger contribution to crop classification than the VIs and GLCM textures. A subset of 60 features was selected using the RF-based feature selection (FS) method, and in this subset, the near-infrared reflectance and the image acquired in August (jointing and heading stages) were found to be the best for crop classification.     

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.     

A systematic comparison of different object-based classification techniques using high spatial resolution imagery in agricultural environments

Geographic Object-Based Image Analysis (GEOBIA) is becoming more prevalent in remote sensing classification, especially for high-resolution imagery. Many supervised classification approaches are applied to objects rather than pixels, and several studies have been conducted to evaluate the performance of such supervised classification techniques in GEOBIA. However, these studies did not systematically investigate all relevant factors affecting the classification (segmentation scale, training set size, feature selection and mixed objects). In this study, statistical methods and visual inspection were used to compare these factors systematically in two agricultural case studies in China. The results indicate that Random Forest (RF) and Support Vector Machines (SVM) are highly suitable for GEOBIA classifications in agricultural areas and confirm the expected general tendency, namely that the overall accuracies decline with increasing segmentation scale. All other investigated methods except for RF and SVM are more prone to obtain a lower accuracy due to the broken objects at fine scales. In contrast to some previous studies, the RF classifiers yielded the best results and the k-nearest neighbor classifier were the worst results, in most cases. Likewise, the RF and Decision Tree classifiers are the most robust with or without feature selection. The results of training sample analyses indicated that the RF and adaboost. M1 possess a superior generalization capability, except when dealing with small training sample sizes. Furthermore, the classification accuracies were directly related to the homogeneity/heterogeneity of the segmented objects for all classifiers. Finally, it was suggested that RF should be considered in most cases for agricultural mapping.     

粗集属性划分的集成遥感分类

提出了一种基于粗集属性划分的遥感分类新方法, 构造了基于粗集的集成遥感分类器。该分类器利用粗集理论将输入的属性集合划分为多个约减, 利用这些约减构造多个训练子集。每个训练子集训练神经网分类器, 在决策时将多个单个分类器的结果进行投票选举。这种方法即减少了单个分类器的输入属性个数, 又避免了由于属性选取造成单一分类器在某些分类上的错误偏见。该分类器与神经网分类器方法, 以及属性选取与神经网结合方法进行了比较。结果表明RSEC无论在分类精度上, 还是在不同样本个数条件下的精度稳定程度上均有较好表现。     

综合多特征的Landsat 8时序遥感图像棉花分类方法

传统的多时相遥感图像分类大多拘泥于单一特征,本文基于多时相的Landsat 8遥感数据,开展了综合多特征的特征提取与特征选择方法研究。综合了NDVI时间序列、最佳时相反射率光谱特征以及纹理特征作为初始分类特征,并采用基于属性重要度的粗糙集特征选择算法对其进行特征约简。分类结果表明:(1)利用初始分类特征,分类的总体精度达到92.81%,棉花提取精度达87.4%,与仅利用NDVI时间序列相比,精度分别提高5.53%和5.05%;(2)利用粗糙集选择后的特征分类,分类总体精度可达93.66%,棉花分类精度达92.73%,与初始分类特征提取结果相比,棉花分类精度提高5.33%。基于属性重要度的粗糙集特征选择不仅提高了分类精度,同时有效降低了分类器的计算复杂度。     

通过训练样本采样处理改善小宗作物遥感识别精度

训练样本质量是决定农作物遥感识别精度的关键因素,虽然高空间分辨率卫星的发展有效地解决了农作物遥感识别过程中的混合像元问题,但是当区域内不同作物种植面积差异较大时,训练集中不同类别样本数量往往相差较大,这样的不均衡数据集影响分类器的训练,导致少数类别的识别精度不理想。为研究作物遥感识别过程中的不均衡样本问题,本文基于GF-2号卫星数据,首先挖掘了地物的光谱信息、纹理信息,用特征递归消除RFE (Recursive Feature Elimination)方法进行特征优选,然后从数据处理的角度采用了5种采样算法对不均衡训练集进行处理,最后使用采样后的均衡数据集训练分类器,对比数据采样前后决策树与Adaboost(Adaptive Boosting)两种分类器的识别结果,发现:(1)经过采样处理后两种分类算法明显提升了小宗作物的分类精度;(2)经过ADASYS (Adaptive synthetic sampling)采样处理后,分类器性能提升最多,决策树的Kappa系数提高了14.32%,Adaboost的Kappa系数提高了10.23%,达到最高值0.9336;(3)过采样的处理效果优于欠采样,过采样对分类器的性能提升更多。综上所述,选择合适的采样方法和分类方法是提高不均衡数据集遥感分类精度的有效途径。     

Selection of Optimal Object Features in Object-Based Image Analysis Using Filter-Based Algorithms

With the increase in spatial resolution of recent sensors, object-based image analysis (OBIA) has gained importance for producing detailed land use maps. One of the main advantages of OBIA is that a variety of spectral, spatial and textural features can be extracted for the segmented image objects that are later utilized in classification. However, using a large number of features not only increases the required computational time, but also requires a large number of ground samples, which is unavailable in most cases. For these reasons, feature selection (FS) has become an important research topic for OBIA based classification studies. In this study, three filter-based FS algorithms namely, Chi square, information gain and ReliefF were applied to determine the most effective object features that ensure high separability among landscape features. For this purpose, importance degree (i.e. ranks) of 110 input object features were firstly estimated by the algorithms, and correlation-based merit function was then applied to determine optimum feature subset size. Multi-resolution segmentation algorithm was applied for segmenting a WorldView-2 image. Support vector machine, random forest and nearest neighbour classifiers were all utilized to classify segmented image objects using the selected object features. Results revealed that the FS algorithms were effective for selecting the most relevant features. Also, the classifiers produced the highest performances with 24 out of 110 features selected by the information gain (IG) algorithm. Particularly, the support vector machine classifier produced the highest overall accuracy (92.00%) with 24 selected features determined by the IG algorithm. A significant improvement of about 4% was achieved by applying FS procedures that was found statistically significant in terms of Wilcoxon signed-ranks test.     

A Robust Texture Analysis and Classification Approach for Urban Land‐Use and Land‐Cover Feature Discrimination

Abstract

Attempts to analyze urban features and to classify land use and land cover directly from high‐resolution satellite data with traditional computer classification techniques have proven to be inefficient for two primary reasons. First, urban landscapes are composed of complex features. Second, traditional classifiers employ spectral information based on single pixel value and ignore a great amount of spatial information. Texture plays an important role in image segmentation and object recognition, as well as in interpretation of images in a variety of applications. This study analyzes urban texture features in multi‐spectral image data. Recent developments in the very powerful mathematical theory of wavelet transforms have received overwhelming attention by image analysts. An evaluation of the ability of wavelet transform in urban feature extraction and classification was performed in this study, with six types of urban land cover features classified. The preliminary results of this research indicate that the accuracy of texture analysis in classifying urban features in fine resolution image data could be significantly improved with the use of wavelet transform approach.     

Utility of image point cloud data towards generating enhanced multitemporal multisensor land cover maps

Multitemporal land cover classification over urban areas is challenging, especially when using heterogeneous data sources with variable quality attributes. A prominent challenge is that classes with similar spectral signatures (such as trees and grass) tend to be confused with one another. In this paper, we evaluate the efficacy of image point cloud (IPC) data combined with suitable Bayesian analysis based time-series rectification techniques to improve the classification accuracy in a multitemporal context. The proposed method uses hidden Markov models (HMMs) to rectify land covers that are initially classified by a random forest (RF) algorithm. This land cover classification method is tested using time series of remote sensing data from a heterogeneous and rapidly changing urban landscape (Kuopio city, Finland) observed from 2006 to 2014. The data consisted of aerial images (5 years), Landsat data (all 9 years) and airborne laser scanning data (1 year). The results of the study demonstrate that the addition of three-dimensional image point cloud data derived from aerial stereo images as predictor variables improved overall classification accuracy, around three percentage points. Additionally, HMM-based post processing reduces significantly the number of spurious year-to-year changes. Using a set of 240 validation points, we estimated that this step improved overall classification accuracy by around 3.0 percentage points, and up to 6 to 10 percentage points for some classes. The overall accuracy of the final product was 91% (kappa = 0.88). Our analysis shows that around 1.9% of the area around Kuopio city, representing a total area of approximately 0.61 km², experienced changes in land cover over the nine years considered.