基于国产高分辨率光学遥感影像的水体提取

遥感图像中地表水体同山体、建筑物等地物产生的阴影在光谱特征上存在较高的类间相似性,导致提取过程中容易出现混淆和错分的情况。针对此问题,提出一种基于面向对象和人工蜂群的地表水体提取方法。该方法首先对遥感图像进行分割以获取分割对象的光谱、比率、几何形状等统计特征,以弥补高分遥感图像波段数目少,信息量不足的缺陷;并借助人工蜂群算法在解决复杂问题最优化方面的优势,选取水体同阴影二值分类的几何平均正确率作为算法的适应度函数,最终获取地表水体的最优化提取规则。选取厦门市大嶝岛和湖南省资兴市部分区域,基于国产高分一号、二号遥感数据进行水体提取,并与传统SVM分类结果进行比较。实验结果表明本算法提取水体的总体精度和Kappa系数均优于传统SVM分类器,表明该方法可应用于高分遥感图像的地表水体提取。     

面向对象的遥感图像分类方法研究

影响遥感图像分类效果的主要因素之一是空间分辨率。通过融合多分辨率遥感图像,引入面向对象的思想,有效地克服了多光谱图像空间分辨率低的问题。该方法由图像分割和分类等一系列技术组成,首先用基于区域分割法则对正射校正SPOT图像进行分割,然后把它作为参考用最大似然法分类器和其他一些经验规则对TM图像进行分类。对土地覆盖图分类进行精度测试,取得了良好的应用效果。     

核概率距离聚类方法及应用

研究聚类分析新方法一直是统计学和机器学习研究领域普遍关注的课题。针对概率距离聚类算法不能解决非线性可分聚类问题的缺欠,笔者应用核函数理论将该模型拓展成为一种能够解决非线性可分聚类问题的统计模型,称为核概率距离聚类分析模型。研制出一种应用新模型进行遥感图像非监督分类研究的实施策略和可行算法;在GDAL遥感图像数据输入输出函数库基础上,用VC++语言开发了遥感图像核概率距离聚类分析算法程序;用ERDAS软件提供的一幅7波段491像素×440像素大小的TM图像进行新方法分类应用实验研究。对比了新模型和其原版本的TM遥感图像非监督分类效果,结果表明新模型的非监督分类效果优于原有的分类模型。     

Variational Bayesian independent component analysis for spectral unmixing in remote sensing image

Spectral unmixing is a key technology of optical remote sensing image analysis; it not only influences the accuracy of the extraction of land cover information and automatic classification of topographical objects, but also greatly hinders the development of quantitative remote sensing. Independent component analysis (ICA) is a statistical method which is recently developed to extract the independent linear components, and which can realize the extraction of endmembers as well as fractional abundances with little a priori knowledge. However, ICA still cannot process the correlations among the various components. To overcome this problem, variational Bayesian independent component analysis (VBICA) has been proposed to process optical remote sensing images. In the Bayesian framework, the separation of independent components of remote sensing image has finally been achieved with conditional independence standards of Bayesian network and approximate variational algorithm. In the simulative image and real AVIRIS hyperspectral remote sensing image, the VBICA algorithm demonstrates its better performance. The experiment’s results indicate that the proposed VBICA algorithm is feasible, which has obvious advantages and a good application prospect. The reason is that it can effectively overcome the correlations between the various components in remote sensing images and break through the limitations of traditional remote sensing images analysis. Last but not least, the VBICA algorithm is applied in the classification of the TM multispectral remote sensing images. Compared to basic maximum likelihood classification, principal component analysis and FastICA algorithms, VBICA improves the classification accuracy of remote sensing images, and contributes to the further extension of the application of ICA in remote sensing image analysis.     

喀斯特地区遥感影像解译新算法——支持向量机算法

现行的遥感影像解译方法有监督分类和非监督分类。在监督分类中有平行算法,最小距离算法、最大似然算法等,而支持向量机是监督分类中的一种新的算法。本研究选择贵阳市花溪区小碧乡局部地区为研究对象,采用SPOT数据,分别运用最大似然算法和支持向量机算法对研究区遥感影像进行解译。通过建立混淆矩阵,来计算分类精度和Kappa系数。结果表明:支持向量机具有分类精度高,分类图斑完整等优点;但在时间的消耗上,支持向量机算法要比最大似然算法长。对于这两种算法而言,都存在地物光谱特征明显相异的地物易于区别,光谱相似的地物容易造成错分的现象,然而支持向量机分类精度要比最大似然分类精度高一些。支持向量机对样本数量具有敏感性,样本数量过多将导致运算时间过长。因此在实际运用中应根据实际情况,选择适合的算法。      

Fine classification of construction land using high-resolution remote sensing images: a case study in planning restricted zone of nuclear power plant

Detailed construction land information plays a significant role in monitoring planning restricted zone of nuclear power plant and ecological environment protection. This study focuses on developing fine classifying method of construction land in planning restricted zone of nuclear power plant using high spatial resolution GF(GaoFen)-1 remote sensing images. The object-oriented classification method is used in this study; the important process of which is image segmentation and classification. Multi-scale segmentation method, rule-based decision tree, and the nearest neighbor classifier are used in classifying construction land classes, i.e., road, industrial, and residential. An optimal segmentation scale is crucial to image segmentation in object-oriented classification. Instead of laborious trial-and-error experiments for optimal image segmentation, the change rates of the local variance in the homogeneous region are calculated to get the optimal segmentation scales. Multi-level classification strategy is used in the following classification. Rule-based decision tree is used to classify road and water, vegetation and non-vegetation, and industrial and residential. And the nearest neighbor classifier is used to classify cropland and forest within the vegetation land use type. The accuracy assessment result shows that the overall accuracy is 89.67% and Kappa coefficient is 0.85 for object-oriented classification, which is much higher than pixel-based maximum likelihood classifier (overall accuracy is 79.17% and Kappa coefficient is 0.74) and support vector machine classifier (overall accuracy is 74.16% and Kappa coefficient is 0.68).     

An adaptive framework for spectral-spatial classification based on a combination of pixel-based and object-based scenarios

Remotely sensed image analysis using spectral-spatial information plays a key role in modern remote sensing applications. This article presents a new semi-automatic framework for spectral-spatial classification of hyperspectral images. The proposed framework benefits from a combination of pixel-based and object-based classification scenarios in which the main parameters are adaptively tuned. In order to reduce the complexity of the method, an unsupervised band selection technique is used as well. Meanwhile, the wavelet thresholding is applied in order to smooth the selected bands. The classification results after applying the proposed method to well-known standard hyperspectral datasets are better than those of the most of the other state-of-the-art approaches. As an example, the overall classification accuracy achieved by applying the proposed semi-automatic spectral-spatial classification framework to the Salinas dataset is more than 99% for 10% training samples per class. Moreover, the vital parameters are adaptively set in our approach.     

基于极限学习机的GF-2影像分类

遥感图像分类是提取图像有效信息过程中重要的一部分,为了探寻最优的分类方法,许多机器学习算法逐步应用于遥感分类中。极限学习机（extreme learning machine,ELM）以其高效、快速和良好的泛化性能在模式识别领域得到广泛应用。本文采用训练速度快、运算量小的极限学习机算法与支持向量机（support vector machines,SVM）算法和最大似然法进行分类对比,对高分辨率遥感图像进行分类,分析极限学习机算法对于遥感图像分类的准确度等性能。选取吉林省长春市部分区域的GF-2遥感数据,将融合后的影像设置为原始数据,利用3种方法进行分类。研究结果表明,极限学习机算法分类图像总体分类精度达到85%以上,kappa系数达到0.718,与其他分类方法相比分类准确度较高,且极限学习机运行时间比支持向量机运行时间约短2 480 s,约为支持向量机运行时间的1/8,因此具有良好的性能和实用价值。     

基于各类特征对齐迁移网络的多时相遥感图像分类

为了在目标域遥感图像不存在标记数据的情况下实现自动分类,论文提出一种基于特征对齐的迁移网络.网络以各类类心对齐和协方差对齐作为迁移策略,全面描述域间各类别之间的对应关系,实现知识迁移.另外,网络采用线性修正单元作为激活函数,能够产生稀疏特征,提高分类效果.该迁移网络能够同时获得对齐的特征和自适应分类器,不需要目标域的标记数据,实现无监督迁移学习.在多时相的Hyperion高光谱遥感图像和WorldView-2多光谱遥感图像上的实验结果证明了该迁移网络的有效性.      

Classify-normalize-classify

Changing atmospheric conditions often result in a data distribution shift in remote sensing images for different dates and locations making it difficult to discriminate between various classes of interest. To alleviate this data shift issue, we introduce a novel supervised classification framework, called Classify-Normalize-Classify (CNC). The proposed scheme uses a two classifier approach where the first classifier performs a rough segmentation of the class of interest (COI) in the input image. Then, the median signal of the estimated COI regions is subtracted from all image pixels values to normalize them. Finally, the second classifier is applied to the normalized image to produce the refined COI segmentation. The proposed methodology was tested to detect deforestation using bitemporal Landsat 8 OLI images over the Amazon rainforest. The CNC framework compared favorably to benchmark masks of the PRODES program and state-of-the-art classifiers run on surface reflectance images provided by USGS.     

Shape analysis and multispectral classification in geological remote sensing

This paper discusses the usage of mathematical morphology in image processing of remotely-sensed data for geologic interpretation. Particular attention is given to noise-reducing transformations of spectral bands before and after different methods of classification, and to the usage of textural context. The development of a viable processing strategy requires a multidisciplinary approach and expert knowledge in different areas: (a) geology, geomorphology, and vegetation in a study area, (b) properties of the sensor for imagery photointerpretation, (c) spectral/spatial properties of the digital data within an integrated dataset (remote sensing and ancillary data), and (d) data-processing tools including mathematical morphology theory. Examples of geometric characterization of Canadian LANDSAT scenes are described in which shape measurements are obtained using a PC-based hybrid image-processing and geographic information system, termed ILWIS, which was developed at ITC, in the Netherlands. Classes from supervised and unsupervised classification are compared to guide in geological mapping. Classes over individual occurrences of broad vegetation-landform units are studied to aid in environmental mapping. Field knowledge is the context necessary to construct expert procedures to drive sequences of data-processing steps toward a target result such as optimal classification, enhancement, or feature extraction. The interaction between expert rules and the image-processing steps can be based on synthetic measurements of shape to quantize the information either spatially or spectrally. Many useful geometrical transformations of spatially-distributed data are extensions or generalizations of spatial analysis functions typical of geographic information systems.     

基于ALOS遥感影像的湿地地表覆被信息提取研究

为了验证ALOS遥感影像湿地地表覆被信息提取的可行性,以黑龙江省三江平原典型内陆淡水沼泽湿地为研究对象,通过ALOS遥感影像波段的光谱及纹理特性分析,探讨适合水体、旱地、水田、沼泽、林地、建设用地、草甸等覆被类型的分类特征;基于非监督、监督及面向对象分类方法,遴选能实现最优分类结果的特征组合,为湿地地表覆盖分类数据源及方法的选择提供参考。结果表明：非监督、监督及面向对象分类方法的总体精度分别达到63.86%、96.14%和85.26%;非监督分类方法整体分类效果不够理想;面向对象方法虽然得到了相对较高的分类精度,但是针对建设用地、林地及草甸地类信息提取的精度处于较低水平;监督分类方法能取得较好效果,最适合于湿地地表覆被信息提取。     

模糊理论在遥感图像分类中的应用

利用2000年假彩色遥感图像,采用模糊C-均值法中的欧氏距离和马氏距离法对崇明东滩的遥感图像进行了处理。通过对白色覆盖物、未耕种土地、一号水稻田、水体和二号水稻田的分类结果表明,欧氏距离的聚类结果优于马氏距离。     

The comparison of PCA and discrete rough set for feature extraction of remote sensing image classification – A case study on rice classification, Taiwan

Texture information offers an extensive solution for image classification by providing better accuracy of image information. However, huge amounts of improper additional texture information may result in a chaotic state, and this leads to uncertainty in the classification process. Considerable portion of earlier works have been carried out through the generally acknowledged procedure of Principal Components Analysis (PCA). However, the PCA method has flaws in the area of influenced and non- influenced attributes. On the whole, whether PCA provides an effective solution to determine the value of knowledge rule in image information still remains a question. This study proposes an innovative method, called Discrete Rough Set method, as a tool for image classification. This study focuses on two crucial issues: (1) The core attributes of the target categories in image classification are systematically analyzed while eliminating surplus attributes rationally; (2) The unique point of each attribute, which influenced the target categories, is successfully found. This is a crucial aspect that is very helpful for the construction of decision rule. Finally, in this study we utilized the expert knowledge classifier and the overall accuracy of Discrete Rough Set (96.67%) exceeds that of the conventional PCA (86.00%) of paddy rice area evaluation from Quickbird image. This result shows that the appropriate classification knowledge can be presented by Discrete Rough Set, and this information can effectively improve the accuracy of image classification. An erratum to this article can be found at     

基于统计学习方法的安徽省安庆市自然资源自动化监测——以山体为例

遥感作为一种可以快速、大范围获取地表覆盖信息的技术手段,为复杂的自然资源调查任务提供了可靠的数据来源。针对山体确界问题,以遥感卫星影像为数据支撑,采用非监督的统计学习方法,为山体特征建模。然后,采用DBSCAN算法和边缘检测思想,识别山体区域,并提取山体边界。该方法不依赖于人工标记真值,实现了山体边界的全自动识别。实验采用安庆市Landsat 8遥感卫星影像数据,有效识别了安庆市境内的山体,并提取山体边界。通过定性和定量化分析,验证了方法的可靠性,证明了遥感技术和统计学习理论在自然资源调查领域的应用潜力。该研究方法和结果能够为安庆市明确山体范围,界定山体的完整性与山体保护规划工作提供理论支撑。     

An unsupervised multi-scale segmentation method based on automated parameterization

The continuous improvement of the launched satellites’ spatial and spectral resolutions has brought new challenges for remote sensing image segmentation technology. The traditional supervised methods greatly depend on artificial interpretation and reduce the degree of automation and robustness of image segmentation. Therefore, the article proposes a novel unsupervised multi-scale segmentation method for high-resolution remote sensing images based on automated parameterization and it mainly includes three steps, adaptive selection of scale parameter (SP) based on local area homogeneity index J-value, multi-scale segmentation based on the inter-scales boundaries constraint strategy, and region merging based on multi-features. The article makes experiments by multi-group high-resolution remote sensing images of different launched satellites and compares the proposed method with the well-known commercial software eCognition and a traditional supervised method. The results show that the proposed method can locate the object edges more accurately and extract the object outlines more completely, and needs no human intervention in segmentation process, so it can provide a generic and effective unsupervised solution for high-resolution remote sensing image segmentation.     

IKONOS遥感数据在离子吸附型稀土矿区环境污染调查中的应用研究——以赣南寻乌地区为例

稀土矿山的开采活动产生了一系列的环境问题。为解决我国南方离子吸附型稀土矿山环境监测的问题,本文选取赣南寻乌地区为研究区,针对目前离子吸附型稀土矿存在的两大环境问题——土地荒漠化及水体污染,采用IKONOS高分辨率遥感数据进行监测方法研究,运用光谱角分类算法提取了研究区土地荒漠化较为严重的区域,运用ISODATA非监督分类算法对稀土矿开采周边河流污染程度进行评估。通过提取结果分析及野外调研,表明高空间分辨率遥感数据处理与分析为离子吸附性稀土矿矿山环境快速、动态监测提供了良好的技术手段。     

不透水面遥感提取及应用研究进展

不透水面信息的提取方法与应用是近年来城市规划、热岛效应分析、水环境监测和水资源管理等诸多领域的研究热点.遥感技术的发展使不透水面快速准确提取成为可能.从影像特征(光谱、空间几何、时间)选择、分类器(参数、非参数)选择和空间尺度(像元、亚像元尺度)选择3个方面归纳和总结了各种不透水面遥感提取方法原理、应用现状和存在问题,回顾了不透水面在城市化监测、人口估计、水环境监测、热岛效应分析、水文气候建模分析等领域的应用,指出了不透水面遥感提取和应用的发展方向.     

深度学习在地质学上的应用

自Hinton等使用基于卷积神经网络的深度学习模型赢得Image Net分类比赛以来,深度学习的研究席卷了各个行业。通过介绍深度学习的历史,探索国内地质行业中深度学习模型的使用情况,并介绍深度学习的基础概念(如神经元、神经网络、监督学习和无监督学习等)以及深度学习基础模型中的2个重要网络:深度信念网络(DBN)和卷积神经网络(CNN)。在此基础上,类比深度学习在医学等相关领域的应用,提出了深度学习在地质上的几点应用:利用深度学习在计算机视觉上表现出的强大能力,可以对遥感图像进行聚类、对岩石样品图像进行分类、对岩石薄片数据进行描述;利用深度学习对原始数据表现出的强大识别能力,处理地质异常数据,从而确定成矿靶区的可能位置;利用深度学习的特点,对地震前的声信号数据进行处理,从而判断出地震发生前的剩余时间。     

利用粗糙集理论进行遥感分类信息提取

以辽宁省双台子河口湿地为研究对象,以Landsat 8和HJ-1-A/HJ-1-B的多时相遥感影像为数据源,根据研究区现状,将研究区分为旱地、芦苇、水田、碱蓬、混合植被、水面、滩涂、居民点、养殖塘九个类型.利用时间序列的归一化植被指数提取植被与非植被的分类阈值,采用粗糙集理论和多时相遥感影像,对植被和非植被分别进行分类规则的获取,建立了研究区决策树分类模型.为了进行精度评价,利用相同的训练点又进行了同样基于像元的最大似然法分类.最后利用混淆矩阵对上述两种方法进行了精度评估,基于粗糙集的决策树分类法与最大似然法总体分类精度分别为93.70%和91.62%,Kappa系数分别为0.92和0.90,两项指标值基于粗糙集理论法均比最大似然法有所提高.这为构建决策树分类模型进行湿地地表分类信息提取提供了一条新的研究思路.