缅甸土地覆被遥感制图和空间格局分析

在“一带一路”倡议框架下,中缅经济走廊逐步从概念转入实质规划建设阶段,了解和掌握缅甸土地覆被的空间格局和分布特征对于合理开发利用资源、制定务实的经济廊道建设规划具有重要的战略意义。利用Landsat-8 OLI遥感影像数据,基于多分类器集成的面向对象迭代分类方法（OIC-MCE）,生产了缅甸2015年30 m分辨率土地覆被产品（MyanmarLC-2015）。采用Google Earth高分辨率影像获取验证样本用于产品精度验证,验证结果表明：MyanmarLC-2015产品的总体分类精度为89.05%,Kappa系数为0.87,各类别的用户精度和制图精度均超过72%,能够准确地反映缅甸土地覆被类型的空间格局。根据产品统计,林地是缅甸面积最大的土地覆被类型,占国土面积56.15%,以常绿阔叶林为主,占林地面积83.57%。耕地面积次之,占国土面积27.01%。地形因子对缅甸土地覆被类型空间分布格局有显著的影响,随着海拔升高,呈现出按如下顺序的垂直地带性特征：森林湿地、水田、旱地、落叶灌木林、落叶阔叶林、常绿灌木林、常绿阔叶林、常绿针叶林。从植被生产力的角度来看,缅甸东部、东北部和东南部植...     

深度学习在GlobeLand30-2010产品分类精度优化中应用研究

本文提出结合深度卷积神经网络与在线高分遥感影像的分类方法，用于GlobeLand30地表覆盖产品的质量优化。首先，通过对多源地表覆盖产品的一致性分析，构建深度学习训练所需的高分辨率遥感大样本（224万样本量）；其次，基于该大规模样本集训练适用于GlobeLand30优化的深度卷积神经网络模型（GoogleNet Inception V3）；最后，利用训练好的神经网络模型对在线高分影像进行分类，用以优化GlobeLand30产品的不可靠区域。经独立测试样本集验证，经过训练的神经网络分类总体精度为87.7%，Kappa系数为0.86，相比原始GlobeLand30的精度（总体精度75.1%、Kappa系数0.71）有了明显提升。在4个试验区的GlobeLand 30产品优化实验表明：该方法能够有效优化GlobeLand30产品的分类精度。     

深度学习在GlobeLand30-2010产品分类精度优化中应用研究

本文提出结合深度卷积神经网络与在线高分遥感影像的分类方法,用于GlobeLand30地表覆盖产品的质量优化。首先,通过对多源地表覆盖产品的一致性分析,构建深度学习训练所需的高分辨率遥感大样本(224万样本量);其次,基于该大规模样本集训练适用于GlobeLand30优化的深度卷积神经网络模型(GoogleNet Inception V3);最后,利用训练好的神经网络模型对在线高分影像进行分类,用以优化GlobeLand30产品的不可靠区域。经独立测试样本集验证,经过训练的神经网络分类总体精度为87.7%,Kappa系数为0.86,相比原始GlobeLand30的精度(总体精度75.1%、Kappa系数0.71)有了明显提升。在4个试验区的GlobeLand 30产品优化实验表明:该方法能够有效优化GlobeLand30产品的分类精度。     

深度学习在GlobeLand30-2010产品分类精度优化中应用研究

本文提出结合深度卷积神经网络与在线高分遥感影像的分类方法,用于GlobeLand30地表覆盖产品的质量优化。首先,通过对多源地表覆盖产品的一致性分析,构建深度学习训练所需的高分辨率遥感大样本（224万样本量）;其次,基于该大规模样本集训练适用于GlobeLand30优化的深度卷积神经网络模型（GoogleNet Inception V3）;最后,利用训练好的神经网络模型对在线高分影像进行分类,用以优化GlobeLand30产品的不可靠区域。经独立测试样本集验证,经过训练的神经网络分类总体精度为87.7%,Kappa系数为0.86,相比原始GlobeLand30的精度（总体精度75.1%、Kappa系数0.71）有了明显提升。在4个试验区的GlobeLand 30产品优化实验表明：该方法能够有效优化GlobeLand30产品的分类精度。     

基于MODIS时序数据的澳大利亚土地利用/覆被分类与验证

选取对气候变化敏感的澳大利亚作为研究区,基于MOD13Q1数据,对澳大利亚2000年土地利用/覆被进行分类。通过Savizky-Golay滤波方法构建高质量NDVI时序数据,为分类奠定数据基础。采用了以决策树为主的混合分类方法对研究区土地利用/覆被进行分类,该方法综合利用了ISODATA分类结果、NDVI阈值及其时间序列主成分分析特征量等数据。通过面积对比和空间位置匹配等多角度验证的方法,综合比较MOD12Q1,GLC_2000与本研究的结果,发现本研究的总体分类精度为63.65%,Kappa系数为0.56,较以上两种已有的土地覆盖产品具有一定优势。     

基于ALOS PALSAR数据时相特征的土地覆盖识别

合成孔径雷达遥感具备全天时、全天候的观测能力,是多时相数据获取的有效保证。以福建省漳浦县为研究区,利用ALOS PALSAR双极化数据开展土地覆盖识别研究。首先基于多时相的强度数据构建时相稳定性指数,基于重复轨道干涉数据的相位信息计算相干性,以此分析和描述该地区典型地物的雷达数据时相特征。然后以典型地物的时相特征为基础,构建决策树分类器,进行土地覆盖识别。最后以实地考察数据、ALOS AVNIR\|2影像和Google Earth影像为参考,进行分类结果的精度评价,总体精度达到81.43%,比利用不同时期的后向散射强度图像为输入波段的最大似然法的分类精度(总体精度为63.06%)高出很多。结果表明:在分类中有效融合时相信息,可以充分提高地物的可分性。     

基于双树复小波分解的Boosting集成学习土地覆被分类研究

近年来,集成学习（Ensemble Learning,EL）分类方法成为土地覆被分类的研究热点,尤其是Boosting集成分类方法具有分类精度高、泛化能力强,在土地覆被分类中得到了显著的应用。但是,Boosting集成分类方法对噪声很敏感,如果训练样本含有噪声时,Boosting算法可能会失效,这是该方法的局限性。为了解决Boosting集成方法在土地覆被分类中存在的问题,有效克服噪声的影响,减少分类结果中的“椒盐”现象和提高分类精度,提出了基于双树复小波分解的Boosting集成学习分类方法。该方法对影像的光谱波段进行一层双树复小波分解,降低图像的噪声,将分解后的各波段作为Boosting集成学习的输入,得到最终的分类结果。实验先后比较了GBDT、XGBoost、LightGBM 3种Boosting集成学习算法在SPOT 6和Sentinel-2A影像上的分类效果。结果表明：(1)在SPOT 6影像上,3种Boosting集成算法总体分类精度均高于90%;DTCWTLightGBM分类总体精度最高,达到94.73%,Kappa系数为0.93,比LightGBM总体精度提高了1.1%...     

基于CNN的吉林一号卫星城市土地覆被制图潜力评估

以吉林一号视频07B星高分遥感影像为基础,采用卷积神经网络（CNN）对城区土地覆被进行精细分类,设置多组光谱变量集合,并与最大似然法、多层感知机和支持向量机分类方法进行对比,全面评估分析各方法对城区土地覆被信息提取的适用性及波谱特征对分类精度的影响。结果表明：CNN模型的分类精度最高,总体精度高于90%,相比其他方法提高幅度达12%以上,能够显著降低“椒盐”噪音;红边波段对所有方法总体分类精度贡献十分有限,而近红外波段对分类精度的提升较为明显;总体而言,红边和近红外波段对CNN分类精度影响较为微弱。深度学习应用于吉林一号高分遥感数据能获取高精度城区土地覆被分类图,可为城市土地资源配置,城市规划与管理提供重要的支撑。     

物候特征辅助下的随机森林宏观尺度土地覆盖分类方法研究

通过遥感技术获取大范围土地覆盖信息对于监测、理解和预测自然资源具有重要的科学意义.MODIS数据是当今宏观尺度土地覆盖研究的主要数据源.本文以河北省为研究区,应用MOD13Q1数据产品,构建MODIS NDVI时间序列,从中反演物候特征作为参与分类的主要辅助信息,并采用随机森林分类方法进行宏观尺度土地覆被分类实验,并与单决策树(CART)进行对比分析.实验结果表明,物候特征辅助下的随机森林宏观尺度土地覆被分类方法的总体精度为87.2％,Kappa系数为0.83,比CART单一决策树精度提高了17.9％;应用物候特征参与分类,使得总体精度提高2.6％;其中,旱地和建筑用地精度分别提高了6.7％和11.9％.     

利用地形参数提高土地覆被分类精度方法的改进

为提高MODIS卫星影像土地覆被产品的分类精度,以京津冀为研究区,在1∶25万土地覆被数据与MODIS土地覆被产品(MCD12Q1)分类一致区内,构建土地覆被类型面积占比与地形因子之间的多元回归模型,并据此改进MODIS土地覆被产品中分类精度较低区域的分类。用面积构成比例和空间一致性比率两个评价指标对改进结果进行评价。结果表明:林地、草地、耕地三种地类的回归模型适合用来改进MODIS土地覆被产品的分类,三种地类与参考数据的空间一致性比率比改进前分别提高了30.02%、40.87%和4.94%;对于与地形因子关系密切的林地和草地,两个评价指标均显示,基于分类一致区建模来改进目标产品的分类精度,比基于整个区域建模改进目标产品的分类精度的效果更加明显。其中,林地的空间一致性比率的提升幅度由8.55%升到30.02%,草地由27.44%升到40.87%。由此可见,地形地貌对土地覆被类型的形成具有重要影响,土地覆被类型面积占比与地形因子之间具有很强的相关关系,基于这种定量关系对土地覆被分类进行改进是完全可行的。     

Rapid Land Cover Classification Using Landsat Time Series based on the Google Earth Engine

Accurate maps of land cover at high spatial resolution are fundamental to many researchs on carbon cycle, climate change monitoring and soil degradation. Google Earth Engine is a cloud-based platform that makes it easy to access high-performance computing resources for processing very large geospatial datasets. It offer opportunities for generating land cover maps designed to meet the increasingly detailed information needs for science,monitoring, and reporting.In this study, we classified the land cover types in Shanxi using Landsat time series data based on the Google Earth Engine Platform. We selected 1 580 sample points be visual interpretation of the original fine spatial resolution images along with Google Earth historical images over six different cover types. We defined training data by randomly sampling 60% of the sample points. The remaining 40% was used for validation. We generated two diffirent types of Landsat composite: (1) one based on median values which is used as the input image for single-date classification; (2)one based on percentile values which is used as input images for time series classification. Random forest classification was performed with two different types of Landsat composites. Random forest classification was performed with two different types of Landsat composites.We visually compared the single-date based to the time series based cover maps of 1990, 2000, 2010 and 2017 in five local areas, and we future compared the results of time series to other products. We aslo performed an accuracy assessment on the land cover classification products. The results shown: (1) The results of time series classification had an overall accuracy of 84%~94%. The time series results improved overall accuracy by 5%~10% compared to single-date results; (2) The result of time series achieves the classification accuracy of products such as CNLUCC, GlobeLand30 and FROM-GLC.The following conclusions were drawn: (1) Cloud computing and archived Landsat data in the GEE has many advantages for land cover classification at a large geographic scale, such as s strong timeliness, short time cycle and low cost; (2) The statistics metrics from Landsat time series is a viable means for discrimination of land cover types, which is particularly useful for the time series classification.     

Finer resolution observation and monitoring of global land cover: first mapping results with Landsat TM and ETM+ data

We have produced the first 30 m resolution global land-cover maps using Landsat Thematic Mapper (TM) and Enhanced Thematic Mapper Plus (ETM+) data. We have classified over 6600 scenes of Landsat TM data after 2006, and over 2300 scenes of Landsat TM and ETM+ data before 2006, all selected from the green season. These images cover most of the world's land surface except Antarctica and Greenland. Most of these images came from the United States Geological Survey in level L1T (orthorectified). Four classifiers that were freely available were employed, including the conventional maximum likelihood classifier (MLC), J4.8 decision tree classifier, Random Forest (RF) classifier and support vector machine (SVM) classifier. A total of 91,433 training samples were collected by traversing each scene and finding the most representative and homogeneous samples. A total of 38,664 test samples were collected at preset, fixed locations based on a globally systematic unaligned sampling strategy. Two software tools, Global Analyst and Global Mapper developed by extending the functionality of Google Earth, were used in developing the training and test sample databases by referencing the Moderate Resolution Imaging Spectroradiometer enhanced vegetation index (MODIS EVI) time series for 2010 and high resolution images from Google Earth. A unique land-cover classification system was developed that can be crosswalked to the existing United Nations Food and Agriculture Organization (FAO) land-cover classification system as well as the International Geosphere-Biosphere Programme (IGBP) system. Using the four classification algorithms, we obtained the initial set of global land-cover maps. The SVM produced the highest overall classification accuracy (OCA) of 64.9% assessed with our test samples, with RF (59.8%), J4.8 (57.9%), and MLC (53.9%) ranked from the second to the fourth. We also estimated the OCAs using a subset of our test samples (8629) each of which represented a homogeneous area greater than 500 m?×?500 m. Using this subset, we found the OCA for the SVM to be 71.5%. As a consistent source for estimating the coverage of global land-cover types in the world, estimation from the test samples shows that only 6.90% of the world is planted for agricultural production. The total area of cropland is 11.51% if unplanted croplands are included. The forests, grasslands, and shrublands cover 28.35%, 13.37%, and 11.49% of the world, respectively. The impervious surface covers only 0.66% of the world. Inland waterbodies, barren lands, and snow and ice cover 3.56%, 16.51%, and 12.81% of the world, respectively.     

A multiple dataset approach for 30-m resolution land cover mapping: a case study of continental Africa

Recent developments in global land-cover mapping have focused on spatial resolution improvement with more heterogeneous features to integrate spatial, spectral and temporal information. In this study, hundreds of features derived from four seasonal Landsat 8 OLI (Operational Land Imager) spectral bands, Moderate Resolution Imaging Spectroradiometer (MODIS) time series vegetation index (VI) data, night-time light (NTL), digital elevation models (DEM) and climatic variables were used for land cover mapping with a target 30-m resolution for the whole African continent. In total, 49,007 training samples (from 11,231 locations) and 23,803 validation samples (from 5,414 locations) interpreted from seasonal Landsat, MODIS Normalized Difference Vegetation Index (NDVI) time series and high-resolution images in Google Earth were used for classifier training (Random Forest) and map validation. Overall accuracy was 76% at 30-m spatial resolution, which is better than previous land cover mapping for the African continent. Besides, accuracies for cropland were improved dramatically by more than 10%. Our method also addressed many remaining issues for 30-m mapping (e.g. boundary effects and declines in resolution). This framework is promising for automatic and efficient global land cover mapping resulting in better visual effects and classification accuracy.     

基于GEE平台的海岛地表覆盖提取及变化监测——以苏拉威西岛为例

随着海洋在国家政治、经济、资源等方面重要性的提升,对海岛开发利用、管理和保护等具有重要意义的海岛地表覆盖研究逐渐受到关注.针对传统的大范围地表覆盖提取速度慢、效率低的问题,基于Google Earth Engine(GEE)地学大数据平台,以作为生态重点保护区的印度尼西亚的苏拉威西岛为研究区,开展海岛地表覆盖提取及变化...     

Evaluation of MODIS Snow Products in Southwestern Xinjiang Using the Google Earth Engine

Google Earth Engine(GEE) is a cloud\|based geospatial processing platform that can analyze geospatial data to achieve parallel processing of massive remote sensing data on a global scale,providing support for remote sensing big data and large\|area research.MODIS snow cover mapping is a global snow cover product established using MODIS data and has been widely used in regional and global climate and environmental monitoring.In the GEE,millions of remote sensing images are stored,including MODIS daily snow products MOD10A1 V5 data and Landsat data.Taking the three research areas in southwestern Xinjiang as examples,the Landsat stored by the GEE were selected,and the NDSI was used to extract the snow cover as the true value of the land cover to evaluate the MOD10A1 accuracy.The results show that the average overall accuracy of MOD10A1 in the snow cover season in southwestern Xinjiang during the period from 2000 to 2016 is 82%,the average misjudgment rate is 2.9%,and the average missed rate is 58.8%.The overall accuracy of MOD10A1 can reach 98% under the clear sky conditions.The accuracy of MOD10A1 is effected by the terrain conditions and cloud cover in different regions.Therefore,the GEE can quickly and effectively filter high quality cloudless Landsat images,and evaluate the accuracy of the MOD10A1 in the snow area around the global regions,displaying intuitively the misjudgment and missed areas in the form of online maps.Meanwhile,GEE provides the Landsat simple cloud score function to calculate the regional cloud cover,which makes the influence of cloud cover on the MOD10A1 accuracy assessment more regionally representative.     

Land Change Patterns and Driving Mechanism in Beijing since 1990 based on GEE Platform

Land-cover and land-use dynamics is a key component for global change,and it is a significant form of the impact of human activities on physical environment.Basing Google Earth Engine platform and Classification And Regression Tree method,selected seven types of cultivated land,forest,grassland,wetland,water body,artificial surface and bare land as classification system,the paper used Landsat 5 TM and Landsat 8 OLI images to interpret the land\|cover and land\|use since 1990 of Beijing.Simultaneously,the paper analyzed and summarized the character of land\|use changing and driving force.The results show that：(1) GEE has outstanding advantages in remote sensing data analysis and processing at regional scales.(2) The CART method has high accuracy of remote sensing classification,and the overall accuracy of validation of 6 land cover products is above 93%.The spatial consistency of 2010 products and GlobeLand30\|2010 data showed that the spatial consistency ratios of woodland,water body and cultivated land were 84.28%,74.75%and 73.56% respectively.The spatial consistency of the distribution is 74.0%.(3) The main land types in Beijing were cultivated land,woodland and artificial surface,and the area accounted for about 90%.During the period from 1990 to 2016,the artificial surface and woodland area increased,and the cultivated land and water were shrinking.The artificial surface area increase of 1 371 km²,and cultivated land shrinkage 40%;On Beijing plain area,artificial surface by the circle of “spread pie” expansion trend to “blossom everywhere” expansion trend;The expansion of the artificial surface is mainly achieved through the encroachment of cultivated land.We constructed a multidimensional stepwise linear equation model to analyze the driving force of land type change,indicated that rapid population growth,rapid economic development,government\|related policies and other socio\|economic development factors jointly drive the Beijing land-cover/land-use evolution process.     

Using long short-term memory recurrent neural network in land cover classification on Landsat and Cropland data layer time series

Land cover maps are significant in assisting agricultural decision making. However, the existing workflow of producing land cover maps is very complicated and the result accuracy is ambiguous. This work builds a long short-term memory (LSTM) recurrent neural network (RNN) model to take advantage of the temporal pattern of crops across image time series to improve the accuracy and reduce the complexity. An end-to-end framework is proposed to train and test the model. Landsat scenes are used as Earth observations, and some field-measured data together with CDL (Cropland Data Layer) datasets are used as reference data. The network is thoroughly trained using state-of-the-art techniques of deep learning. Finally, we tested the network on multiple Landsat scenes to produce five-class and all-class land cover maps. The maps are visualized and compared with ground truth, CDL, and the results of SegNet CNN (convolutional neural network). The results show a satisfactory overall accuracy (> 97% for five-class and > 88% for all-class) and validate the feasibility of the proposed method. This study paves a promising way for using LSTM RNN in the classification of remote sensing image time series.     

Monitoring rubber plantation distribution on Hainan Island using Landsat OLI imagery

Rubber expansion has been widely explored at various scales due to its increasing environmental and socio-economic impact and the global demand for natural rubber products. However, accurate and up-to-date maps and large-scale monitoring of rubber plantation spatial distribution are not yet available. In this article, we developed a simple algorithm for rapidly and accurately mapping rubber plantations on Hainan Island, China, by combining survey samples and Landsat 8 Operational Land Imager data from 2014 in order to understand their spatial distribution. The results showed that rubber plantations are distinguishable from other land-cover types by band value changes, vegetation index changes, and phenological phase changes (defoliation and foliation) and can be accurately extracted from multi-temporal Landsat images using a decision tree method and Google Earth Engine. This method results in a high overall classification accuracy of 92.17% with a corresponding κ of 84.33%. Rubber plantations are concentrated in the northwest of Hainan Island and gradually decrease to the south and east. Most plantations are found in the 50–500-m elevation range, with few found outside this range. We believe that the proposed approach will have significant implications for mapping and monitoring rubber spatial distribution at regional scales.     

Land use/land cover changes in the coastal zone of Ban Don Bay,Thailand using Landsat 5 TM data

Land use/land cover of the Earth is changing dramatically because of human activities and natural disasters. Information about changes is useful for updating land use/land cover maps for planning and management of natural resources. Several methods for land use/land cover change detection using time series Landsat imagery data were employed and discussed. Landsat 5 TM colour composites of 1990, 1993, 1996 and 1999 were employed for locating training samples for supervised classification in the coastal areas of Ban Don Bay, Surat Thani, Thailand. This study illustrated an increasing trend of shrimp farms, forest/mangrove and urban areas with a decreasing trend of agricultural and wasteland areas. Land use changes from one category to others have been clearly represented by the NDVI composite images, which were found suitable for delineating the development of shrimp farms and land use changes in Ban Don Bay.