Landsat-8影像的LDA模型变化检测

变化检测一直是遥感研究领域的热点,随着遥感技术的不断发展,新型数据源不断涌现,使传统遥感变化检测方法面临新的挑战。本文以Landsat-8影像为主要数据源,使用影像分割算法,设计2期遥感影像的文档-单词映射,将影像中所有的像元作为视觉单词,利用LDA模型将影像文档从单词空间转换到主题空间进行表达。在此基础上,结合实地调查对变化区域进行检测和验证,形成一套面向对象的LDA模型变化检测方法。研究表明：基于图斑的分析可有效消除以像元尺度进行变化检测产生的椒盐现象;利用LDA模型构建的变化检测方法能较好地实现影像文档特征的统一表达,有效去除2期影像相同地物因光谱差异导致的变化误检验;与差值法和波谱角等常规遥感变化检测方法相比,该方法能有效地减少错漏判,提高遥感影像变化检测的正确率,为中高分辨率遥感影像的变化检测提供新思路。     

一种基于面积对称模型的土地利用/覆被真伪变化图斑检测方法

不同时相土地利用/覆被数据间的空间配准误差,是产生土地利用/覆被伪变化图斑的一个主要原因。本文从土地利用/覆被原始图斑与其相邻的变化图斑间的空间关系角度,提出了由同一原始图斑产生的土地利用/覆被伪变化图斑的面积对称理论,设计和实现了针对因空间配准误差而导致的土地利用/覆被伪变化图斑自动化检测模型,并以内蒙古自治区通辽市奈曼旗1980年和2000年两期土地利用/覆被图对该模型进行了实验模拟。结果表明：当两期数据的空间配准误差不超过原始影像1个像元时,总体检测精度达到90%以上;误差不超过5个像元时,总体检测精度达到80%以上;误差不超过原始影像10个像元时,总体检测精度达到70％以上。同时,由于运行时唯一需设定的面积对称系数阈值可设为0.2-0.4（默认设为0.3）,该检测模型可适用于由空间配准误差引起的伪变化图斑的自动检查,可满足由于空间配准误差所引起土地利用/覆被伪变化图斑剔除的需求。     

多时相耕地覆盖提取和变化分析：一种结合遥感和空间统计的时空上下文方法

基于多时相影像的耕地提取和变化分析是有效管理和保护耕地资源的重要手段。然而就多时相耕地的分类提取而言,现有方法对于多时相影像中地物的时空特征表达和时空上下文关系建模存在着局限性,导致耕地的提取精度不佳;其次,对于耕地的变化分析,现有方法往往只关注基于行政单元的耕地面积统计变化,而对耕地变化在空间上的相关性分布特点考虑较少。因此,本文首先提出了一种时空上下文分类方法,综合表达和利用多时相影像中地物的光谱、纹理和空间等特征,建模时空维度上地物间在特征和语义上的上下文关系,来提高耕地覆盖分类的精度;其次,基于耕地覆盖的提取结果,在规则格网和行政区划单元上,采用GIS空间统计方法分析耕地变化的空间相关性特点;最后,以北京市顺义区为例,以2015—2019年的多时相Sentinel-2影像为数据源对本文方法进行验证。结果表明,与常见的2种多时相影像分类方法相比,本文方法在多时相耕地分类上精度最高,平均用户精度和制图精度分别达到91.21%和90.53%,所有类别的总体精度为90.79%。这表明本文方法能精确提取多时相耕地覆盖信息。通过对耕地变化的空间分布特点进行分析,发现2015—2019年顺义...     

DMSP/OLS稳定夜间灯光影像在中国的分类校正方法

长时间序列DMSP/OLS夜间稳定灯光影像数据集存在饱和像元大量积聚在城市中心区域及影像之间缺少可比性的问题,导致数据集在进行空间分析及时间对比研究时会受到限制。为解决上述2个问题,本文在现有研究的基础上提出了一种基于不变目标区域的中国稳定灯光影像分类校正方法。该方法对数据集中每一期中国区域稳定灯光影像进行了饱和及非饱和像元的分类。对饱和像元,选择北京等5个城市作为不变目标区域,选择不存在饱和像元的辐射校准的灯光影像作为参考,基于不变目标区域的非饱和像元之间的相关关系也同样适用于饱和像元的假设,对饱和像元进行了校正;同时,基于多期辐射校准的灯光影像之间存在的相互校准关系,对饱和校正后的饱和像元进行了相互校准。对于非饱和像元,选择了长沙等13个城市作为不变目标区域;根据稳定灯光影像中非饱和像元DN值总和随时间变化的总体趋势,确定了13期稳定灯光影像作为参考,其他多期影像基于不变目标区域中非饱和像元之间的相关关系进行了相互校准。为验证校正结果的准确性,多种相互校准和饱和校正方法被引入进行比较。本文提出的分类校正方法与另外2种相互校准方法相比对F10、F12和F14卫星影像校准效果更好,另外2种方法对F15和F16卫星影像的校准结果更好,3种方法总体上都实现了对数据集影像的相互校准且使校正后的影像具有了可比性。与经过植被指数校准后的城市夜间灯光指数（The Vegetation Adjusted NTL Urban Index,VANUI）相比,分类校正方法解决了饱和像元聚集于城市中心区域的问题,减弱了像元饱和程度,校正结果更接近辐射校准的灯光影像。此外,在中国城市水平上,分类校正的稳定灯光影像与GDP和电力消耗值的相关关系与其他2种相互校准方法相比也是相对较好的,能更客观合理地反映区域经济发展的差异。     

Remote sensing classification of western Sierra Leone using landsat TM and ETM +

The study examines the changes of land cover/use resources for the period under investigation.An unsupervised vegetation classification is being performed that provides five distinctive classes and thus assesses these changes in five broad land cover classes-high/moist forests,forest regrowth,mixed savanna,bare land/ grass and water.The remote sensing images used in this work are both images of TM and ETM+in different time periods(1986 to 2001)to determine land cover/use changes.A fairly accuracy report is recorded after performing the unsupervised classification,which shows vegetation has been depleted for over the years.Changes created are mostly human and to a lesser extent environment.Human activities are mainly encroachment thus altering the landscape through activities such as population growth,agriculture,settlements,etc.and environment due to some perceive climatic changes.This vegetation classification highlights the importance to acquire and publish information about the country’s partial vegetation cover and vegetation change including vegetation maps and other basic vegetation influencing factors,leading to an understanding of its evolution for a period.     

基于图分割的高分辨率遥感影像建筑物变化检测研究

建筑物是城市地理数据库中最容易发生变化和最需要更新的部分,其更新工作量巨大,因此开展对高分辨率遥感影像中的建筑物进行自动提取和变化检测研究具有重要的意义.本文以精确提取变化建筑物的位置和轮廓为目标,基于图分割提出一种高分辨率遥感影像建筑物变化检测方法.首先,将遥感影像中的每个像元映射成图的顶点,利用像元之间的距离阈值构造图的边,综合利用位置,灰度和边缘3种特征计算边的权值,将遥感影像的分割转化为图的分割,并用归一化图分割方法得到分割对象集合;然后,以长宽比和矩形度作为约束条件,对2期遥感影像中的分割对象集合进行筛选,提取建筑物对象;最后,根据2期影像中建筑物之间的空间,面积和格局关系识别建筑物的变化类型(包括新增,消失和改建),并对其进行可视化表达.为了验证本文方法的有效性,分别以深圳市的WorldView影像和北京市的QuickBird全色影像为数据源,从中选取13组具有代表性的子影像进行实验.结果表明,本文提出的方法对配准精度较低的影像组具有一定的适应性,容许的配准误差达到20个像元(10 m),平均查准率和平均查全率分别达到93.16%和87.90%.     

土地覆盖的图像融合动态监测

近20年来,关于图像融合应用分析方法,如HIS,PCS,HPF,SFIM,SVR,Wavelet和Brovey等均有新的进展。本文主要是对不同时相的影像进行融合,如Brovey-融合法将不同时相的TM(1986年7月26日)和ETM+(2000年5月4日)的PAN波段影像进行融合,然后对其采用非监督分类和PCS分析,将两时相的土地覆盖变化区域提取出来。同时将两时相影像,用后分类法进行分类提取出变化区域。研究表明融合法具有快速、简便和准确的特点。     

面向地理国情监测的变化检测与地表覆盖信息更新方法

常态化地理国情监测能够全面、动态地掌握地理国情信息及其变化,为经济建设和社会发展提供数据基础。地理国情普查成果是按照统一规范标准、经过内业解译和外业核查形成的矢量数据。如何在普查成果的基础上,利用多时相遥感影像实现变化信息提取与更新是地理国情监测的关键。针对地理国情普查成果的特点与监测需求,以多时相遥感影像处理分析为基础,构建了针对地理国情监测的变化检测方法体系,提出了像元—对象结合的多时相影像变化检测、基于对象实体统计分析的变化识别方法,实现了综合地理国情普查成果和遥感影像的地理国情变化检测与数据更新。基于像元—对象结合的多时相影像变化检测首先根据传统的变化矢量分析法提取基于像元的变化检测结果,再以地理国情普查的矢量对象为统计单元计算对象内变化像元的比例,以此判断该矢量对象是否发生了变化,并根据变化像元的比例计算其变化强度。基于对象实体统计分析的变化检测方法直接以地理国情矢量为对象进行特征提取和差异构造,再将差异影像进行阈值分割得到基于地理国情对象的变化检测图。最后,根据变化检测结果,对变化区域进行面向对象分割,并从上一期未变化区域选取训练样本训练分类器模型以得到变化区域的地表覆盖类型,将变化区域与未变化区域结合得到更新后的地理国情矢量图。选取江阴市地理国情普查成果和两期高分辨率遥感影像进行试验,结果表明本文提出的方法在准确提取和解释变化区域的同时,明显提高了变化检测和数据更新的效率,可用于常态化地理国情监测。     

THE DRIVING FORCES OF LAND USE/COVER CHANGE INTHE UPSTREAM AREA OF THE NENJIANG RIVER

IINTRODUCTION niques in China（ZHANG。t al，1999； CHEN et al，2000b）．Since the early 1990s，he acceleration of land With its abundant forest，land and water re－use／cover change（LUL乙）has spurred renewed con－sources，the area of the Nenjlang River valley Is one ofcerns about the role ofland use change Indrlvingmany the lmpoFtant lumber products and commodity grainenvironmental problems．Research on the causes and ba…     

Fusion and classification of Beijing-1 small satellite remote sensing image for land cover monitoring in mining area

In order to promote the application of Beijing-1 small satellite(BJ-1) remote sensing data,the multispectral and panchromatic images captured by BJ-1 were used for land cover classification in Pangzhuang Coal Mining.An improved Intensity-Hue-Saturation(IHS) fusion algorithm is proposed to fuse panchromatic and multispectral images,in which intensity component and panchromatic image are combined using the weights determined by edge pixels in the panchromatic image identified by grey absolute correlation degree.This improved IHS fusion algorithm outper-forms traditional IHS fusion method to a certain extent,evidenced by its ability in preserving spectral information and enhancing spatial details.Dempster-Shafer(D-S) evidence theory was adopted to combine the outputs of three member classifiers to generate the final classification map with higher accuracy than that by any individual classifier.Based on this study,we conclude that Beijing-1 small satellite remote sensing images are useful to monitor and analyze land cover change and ecological environment degradation in mining areas,and the proposed fusion algorithms at data and decision levels can integrate the advantages of multi-resolution images and multiple classifiers,improve the overall accuracy and produce a more reliable land cover map.     

高分遥感图像相对辐射校正中的伪不变地物自动提取和优化选择

相对辐射校正是遥感变化检测中重要的预处理过程,伪不变地物(Pseudo-Invariant Features,PIF)是多时相影像中相对不变的地物,是相对辐射校正中的重要依据.针对高分遥感图像变化检测中相对辐射校正的要求,本文提出了一个自动提取和优化选择PIF的流程和方法:首先计算两期图像的亮度、光谱特征和空间特征的变...     

西辽河的老哈河流域土地利用遥感动态监测

本文利用Landsat TM和ETM+遥感影像,基于知识规则的遥感分类及分类后处理方法对20世纪80、90年代末的半干旱的老哈河流域土地利用变化进行动态监测,并用GIS方法对土地利用变化及其空间分布进行了定量分析。结果表明:在这十年间,由于气候条件的变化及人类活动的影响,老哈河流域土地利用发生显著变化,变化范围占流域总面积的33.64%,土地利用变化主要表现为农、林、牧用地之间的转换。其中水体、滩地、林地、草地和未利用土地减少,大量的沙地和裸土地被开发利用;城乡用地和旱地有所增多,水田大幅度增加,农业耕作方式发生巨大变化。本研究为研究北方干旱化形成机理的生态-水文耦合模型提供重要的下垫面输入条件,为以后土地利用格局的改变对蒸散发和径流过程的影响提供重要的分析前提。     

基于地学知识的大尺度土地利用/土地覆盖精细化分类方法研究

人类活动对生态环境具有显著影响,大尺度土地利用/覆盖变化（Land Use/Cover Change,LUCC）作为人类活动最直接的表征,能够很好地反映这一过程,因此进行精确而迅速的大尺度土地利用/覆盖分类与提取方法研究尤为关键。全球覆盖产品GlobCover（2005/2006）数据已经具有良好的空间精度和数据准确度,但仍然存在一些分类误差。为提高地表覆被分类精度,本文以GlobCover（2005/2006）的巴西数据为例,以2005年Landsat TM/ETM影像为主要信息源,结合相应地学知识与辅助数据,利用人机交互逐栅格修改方法得到2005年土地利用数据产品。结果表明：通过对GlobCover数据和本次成果数据进行精度评价与对比分析,GlobCover数据巴西地区的总体精度为67.17%,Kappa系数为0.58,改进后产品总体精度为93.39%,Kappa系数为0.91。此外,改进后数据显示巴西常绿阔叶林面积最大,面积比例达45.67%;农地/自然植被镶嵌面积次之,比例为19.19%;封闭灌丛面积最小,比例为12.34%。农地/自然植被镶嵌和灌丛与草地2种地类的修改比例最大,其中混合像元地类比例减少3.54%,灌丛与草地比例增加3.81%。综上,改进方法可以有效地提高土地利用/覆盖分类的效率和精度,为后续大尺度LUCC产品的制作和以LUCC产品为基础的相关研究提供参考。     

FARMLAND AND URBAN AREA DYNAMICS MONITORING INCHINA USING REMOTE SENSING AND SPATIALSTATISTICS METHODOLOGY

1 INTRODUCTIONGlobal change research involves much geo-objectsand geo-process, such as climate and environmentalchange, substance and energy cycling, land-use/land-cover change (LUCC), interactivity between human and nature, etc.. So it need cooperation frommany research communities including international research programs groups such as IGBP (InternationalGeosphere-Biosphere Programme IPCC (Intergovemmental Panel on Climate Change), IHDP (InternationalHuman Dimension Program o…     

历史解译知识引导下组合遥感图谱特征的变化检测方法

高空间分辨率遥感影像为地表变化监测提供了大量细节信息,这使得基于高分辨率影像的变化检测技术成为当前遥感领域的研究热点之一。本文提出了一种历史解译知识引导下组合遥感图谱特征的变化检测方法。首先,通过分割前后时相的组合影像构建空间位置一致的对象,并在提取对象光谱和纹理特征后,引入前期土地覆盖专题图指导2类图谱特征相似度的DS证据融合;然后,利用其历史存档图斑所属区域的优势地类标签指示不同特征相似度的证据差异融合;最后,基于GMM（Gaussian Mixture Mode）模型的二值化方法提取最终的变化区域。实验结果表明,该方法能充分利用历史解译知识指导不同时相高分辨率影像对象特征相似度组合,一定程度上提高了变化检测正确率。     

AN ASSESSMENT OF LAND USE CHANGES IN FUQING COUNTY OF CHINA USING REMOTE SENSING TECHNOLOGY

Fuqing County of southeast Chian has witnessed significant land use changes during the last decade.Re-mote sensing technology using multitemporal Landsat TM images was used to characterize land use types and to monitor land use changes in the county.Two TM scenes from 1991and 1996 were used to cover the county and a five-year time period.Digital image processing was carried out for the remotely sensed data to produce classified images.The images were further processed using GIS software to generate GIS databases so that the data could be further spatially analyzed taking the advantages of the software.Land use change areas were determined by using the change detection technique.The comparison of the two classified TM images using the above technologies reveals that during the five study years,a large area of arable lands in the county has been lost and deforestation has taken place largely because of the dramatic in-crease in built-up land and orchard.The conclusive statistical information is useful to understand the processes,causes and impacts of the land use changes in the county.The major driving force to the land use changes in the county ap-peared to be the rapid economic development.The decision makers of the county have to pay more attention to the land use changes for the county‘s sustainable development.     

High resolution land cover datasets integration and application based on Landsat and Globcover data from 1975 to 2010 in Siberia

Land cover is recognized as one of the fundamental terrestrial datasets required in land system change and other ecosystem related researches across the globe. The regional differentiation and spatial-temporal variation of land cover has significant impact on regional natural environment and socio-economic sustainable development. Under this context, we reconstructed the history land cover data in Siberia to provide a comparable datasets to the land cover datasets in China and abroad. In this paper, the European Space Agency(ESA) Global Land Cover Map(GlobCover), Landsat Thematic Mapper(TM), Enhanced Thematic Mapper(ETM), Multispectral Scanner(MSS) images, Google Earth images and other additional data were used to produce the land cover datasets in 1975 and 2010 in Siberia. Data evaluation show that the total user′s accuracy of land cover data in 2010 was 86.96%, which was higher than ESA GlobCover data in Siberia. The analysis on the land cover changes found that there were no big land cover changes in Siberia from 1975 to 2010 with only a few conversions between different natural forest types. The mainly changes are the conversion from deciduous needleleaf forest to deciduous broadleaf forest, deciduous needleleaf forest to mixed forest, savannas to deciduous needleleaf forest etc., indicating that the dominant driving factor of land cover changes in Siberia was natural element rather than human activities at some extent, which was very different from China. However, our purpose was not just to produce the land cover datasets at two time period or explore the driving factors of land cover changes in Siberia, we also paid attention on the significance and application of the datasets in various fields such as global climate change, geopolitics, cross-border cooperation and so on.     

Landscape Dynamics and Driving Factors in Da＇an County of Jilin Province in Northeast China During 1956-2000

The landscape pattern of Da＇an County, Northeast China has undergone significant changes since the 1950s as a result of climatic change and human activities. The aim of this paper is to quantitatively study landscape pattern and its spatial dynamics of Da＇an County at the landscape level over the nearly 50-year span. Patch dynamics were examined according to land use and land cover change processes built from a series of images, as well as topographic maps, and temporal patterns built from landscape pattern metrics. The transition matrix of landscape patch types and changes of various landscape metrics were applied. The results showed that, from 1956 to 2000, the landscape within the study area had undertaken a complicated evolution in landscape structure and composition. The outstanding characteristic is that saline-alkali land increased and grassland decreased. As some smaller patches amalgamated, the heterogeneity of patch decreased. All those changes were the synthetic result of both climatic and anthropogenic influences, but the predominant factor was different in different parts. In the southern part of the study area, the landscape pattern changes resulted from the modification of climate obviously, while in the northern part, the landscape pattern changes were mainly caused by human activities, such as the conversion between farmland and saline-alkali land. This phenomenon showed that human activities played more important role in the north than in the south of Da＇an County.     

批量变更技术在年度土地变更调查中的应用

从东莞市年度变更调查工作入手,分析变更调查数据的收集、整理、检查、处理流程,结合GIS技术,数据库技术,设计土地利用数据库批量变更技术。最后得出批量变更技术相对于单个图斑的变更在工作流程上和工作效率上的优势。     

基于超像元词包特征和主动学习的高分遥感影像变化检测

为解决高分辨率遥感影像变化检测中存在底层特征缺乏语义信息、像元级的检测结果存在“椒盐”现象以及监督分类中样本标注自动化程度较低,本文提出一种基于超像元词包特征和主动学习的变化检测方法。首先采用熵率分割算法获取叠加影像的超像元对象;其次提取两期影像像元点对间的邻近相关影像特征（相关度、斜率和截距）和顾及邻域的纹理变化强度特征（均值、方差、同质性和相异性）,经线性组合作为像元点对的底层特征;然后基于像元点对底层特征利用BOW模型构建超像元词包特征,并采用一种改进标注策略的主动学习方法从无标记样本池中优选信息量较大的样本,且自动标注样本类别;最后训练分类器模型完成变化检测。通过选用2组不同地区的GF-2影像和Worldview-Ⅱ影像作为数据源进行实验,实验结果中2组数据集的F1分数分别为0.8714、0.8554,正确率分别为0.9148、0.9022,漏检率分别为0.1681、0.1868,误检率分别为0.0852、0.0978。结果表明,该法能有效识别变化区域、提高变化检测精度。此外,传统主动学习方法与改进标注策略的主动学习方法的学习曲线对比显示,改进的标注策略可在较低精度损失下,有效提高样本标注自动化程度。