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。     

影像的土地覆被快速分类

精确的土地覆盖信息是进行碳循环、气候变化监测、土壤退化等相关科学研究的基础。随着云计算技术的不断成熟,一些高效算法与平台被不断提出,用来充分挖掘遥感数据所包含的海量信息。基于Google Earth Engine(GEE)云平台,利用随机森林监督分类法对1990、2000、2010、2017年的山西省土地覆被进行了分类。参考Google Earth高清影像选择的1580个样本点,对分类结果进行了验证;同时将分类结果与CNLUCC、GlobeLand30、FROM-GLC等现有土地覆被分类产品进行比较。验证和对比发现时间序列分类结果的总体精度达到86%~94%,比同期单时相分类总体精度提高了5%~10%;本文时间序列结果达到了CNLUCC、GlobeLand30、FROM-GLC等产品的分类精度。结果表明:①在快速准确土地覆被分类方面,时间序列影像与云平台结合,显示出时效性强、时间周期短、成本低等优势;②时间序列百分位数指标能有效地区分不同土地覆被类型的物候差别,在进行土地覆被分类方面显示出简单、易用、高效等特点。该方法对于深入研究大区域尺度的土地覆被变化过程具有重要的参考价值。     

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。     

中国地区AVHRR长时间序列积雪范围产品精度评估

积雪是冰冻圈重要的组成部分,积雪面积在地表的积累和消融影响着水资源的平衡以及生态环境的变化,进而影响社会经济的发展。目前由日本宇航局发布的JASMES系列产品包含一套北半球自1978年以来长时间序列的AVHRR逐日积雪范围产品,可以作为气候和水文模型的重要输入,然而该产品在中国地区的精度尚未被评估。以高分辨率Landsat-5TM数据和气象台站雪深观测数据两种手段通过构建误差矩阵和一致性检验方法来评价JASMES积雪范围产品在我国三大积雪区(北疆,东北和青藏高原)的总体精度。研究结果表明:使用Landsat-5TM积雪面积二值图进行验证时,东北、北疆和青藏高原地区的总体精度分别为66.3%、69.3%和49.5%,积雪识别的漏分误差均较为严重,混合像元以及积雪消融时间过短严重地降低了青藏高原地区的总体评估精度;在积雪季逐月分析中,JASMES积雪产品由于地形的影响,青藏高原地区总体评估精度低于北疆和东北地区,消融过快的特性进一步降低了评估的总体精度。结合气象站点数据验证时,当雪深阈值为9cm时,青藏高原、北疆和东北地区总体精度分别为94.2%、67.8%和77.8%,青藏高原地区对非雪像元的识别精度达到98.9%;随着站点积雪覆盖日数的增加,其总体精度均呈下降的趋势,漏分积雪情况较为严重,存在多分陆地,少分积雪的现象。综上所述,JASMES积雪产品在青藏高原、北疆和东北地区均存在漏分积雪比较严重的现象,青藏高原地区积雪漏分现象最为突出。     

欧亚大陆中高纬度区逐日无云积雪产品研发及验证

利用多源遥感数据,结合光学遥感数据高空间分辨率及被动微波数据不受云干扰的优势,利用MODIS逐日积雪标准产品和AMSR-E雪水当量产品,生成了欧亚大陆中高纬度区500m分辨率的逐日无云积雪产品,并利用更高分辨率的Landsat-TM数据生成的积雪产品作为"真值"影像,对研发的逐日无云积雪覆盖产品的精度进行了验证。结果表明:MOD10A1和MYD10A1受云影响均较为严重,无法直接用于地表积雪面积的监测。而本研究合成的逐日无云产品具有较好的精度,与TM积雪图具有较高的一致性。但不同的土地覆盖类型对积雪分类精度有一定的影响。其中,裸地和草原覆盖区精度最好,Kappa系数分别为0.655和0.644,均为高度一致性;其次精度较好的是灌丛和耕地覆盖区,Kappa系数分别为0.584和0.572,均为中等的一致性;而森林覆盖区由于受到高大植被的影响,Kappa系数仅为0.389,合成产品相对TM积雪产品明显高估了森林区积雪面积。整体Kappa均值达到0.569,接近高度一致,研究结果对实时监测欧亚大陆积雪面积具有一定的应用价值。     

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...     

利用被动微波遥感数据反演我国积雪深度及其精度评

考虑到我国西部地区使用SSM/I全球算法将高估积雪深度,故以东经105°为界将我国分为东部和西部。在西部地区采用修正后的雪深算法,东部地区沿用全球算法。对散射系数较高,容易和积雪相混淆的降雨、寒漠和冻土地表类型,通过积雪分类树进行剔除,进而发展了一套适用于全国积雪深度的业务化反演方案。最后利用MODIS积雪产品对冬季90天的结果进行了精度评价,总体精度平均达到86.4%,最高精度达到95.5%,Kappa系数均值为65.5%,最大值达到86.2%。     

2000~2010年吉林省积雪时空变化特征及其与气候的关系

利用Terra卫星提供的2000年10月1日到2010年4月30日每日雪覆盖产品MOD10A1,提取研究区积雪覆盖指数SCI、积雪日数SCD、积雪初日SCOD及积雪终日SCMD遥感信息,结合同期吉林省界内23个地面气象观测站的同期气温和降水资料,分析该区积雪的变化特征与气温和降水的关系。结果表明:① 吉林省大部分地区积雪日数为30~90 d,东部山区积雪持续时间长、积雪初日日期早以及积雪终日日期晚,中西部地区变化情况相反;② 积雪覆盖指数SCI呈波浪式变化,与积雪季气温呈负相关;③ 积雪日数与气温呈反相关、与降水量呈正相关,与积雪季气温、夏季降水量的相关系数分别为-0.7407、0.6875;积雪初日情况相反,与积雪季气温、夏季平均气温为0.743、0.5479;积雪终日与气温呈反相关、与降水量呈正相关,与积雪季气温、夏季降水量为-0.5214、0.4647。积雪指数均对气温的变化更敏感,气温升高导致积雪初日推迟、积雪终日提前,从而使积雪日数减小;积雪季降水量的增加有利于积雪日数增大,而积雪终日的推迟有利于夏季降水量的增加。     

全球土地覆盖产品中森林类型数据在中国区域的质量评估

全球土地覆盖产品是森林类型数据的重要来源，系统评估不同产品中森林类型数据质量对数据使用者及生产新的数据具有指导意义。选取CCI-LC、MCD12Q1、Globeland30、GLCFCS30、FROM-GLC10、Esri10和ESA10七套土地覆盖产品，从自身稳定性、面积比较、空间一致性、精度估算4个方面评估其森林类型数据质量。结果显示CCI-LC中森林类型数据稳定性显著高于其他产品，2001～2019年19期数据的稳定性大于94%；就森林总面积而言，Globeland30、GLC-FCS30和Esri10与第九次全国森资源调查数据最为接近，Esri10的灌木存在严重高估现象；Esri10和MCD12Q1与其他产品的空间一致性均较低，ESA10和FROM-GLC10的一致性最高（85.3%）；高空间分辨率产品的总体精度优于中低空间分辨率产品，不考虑灌木时，ESA10、FROM-GLC10和Esri10的总体精度分别为90.63、87.99、85.22;GLC-FCS30、CCI-LC和MCD12Q1中精细森林类型总体精度均低于48%。     

Seasonal Accuracy Assessments of Three Land Cover Datasets in Pakistan

Global land cover datasets play an important role in the fields of ecology, climate and resources. GlobeLand30-2010 (30 m), FROM-GLC-2010 (30 m) and GlobCover-2009 (300 m) are three global high-precision land cover datasets with a wide range of applications. In order to judge whether these data sets are sufficient to describe the real situation of land cover in different seasons, the inter-seasonal accuracy of the aforementioned three global land cover datasets using Pakistan as a representative study area was evaluated. A total of 1 000 land cover sample points were selected from 122 Landsat-5, Landsat-7 multi-spectral remote sensing images during 2009~2011 to generate summer and winter land cover classifications.The results show that the accuracies of summer and winter land cover classifications are different in Pakistan. The overall land cover classification accuracies of GlobeLand30-2010 (65.6% vs. 63.9%) and FROM-GLC-2010 (61.2% vs. 59.0%) in summer are slightly higher than those in winter. The overall accuracy of GlobCover-2009 (59.5% vs. 59.1%) in winter is slightly higher than that in summer. GlobeLand30-2010 performs best in classifying cropland, impervious surface, and water body, FROM-GLC-2010 performs best in classifying vegetation, glaciers, and snow, and GlobCover-2009 performs best in classifying bare land. The classification of cropland, bare land, glaciers, and snow in the three datasets is more in line with the real situation in winter than in summer; the classification of vegetation and water bodies is more in line with the real situation in summer; there is no obvious seasonal difference in impervious surface. There should be at least one sample point per 1 000 square kilometers.     

Exploring point-of-interest data from social media for artificial surface validation with decision trees

Artificial surfaces represent one of the key land cover types, and validation is an indispensable component of land cover mapping that ensures data quality. Traditionally, validation has been carried out by confronting the produced land cover map with reference data, which is collected through field surveys or image interpretation. However, this approach has limitations, including high costs in terms of money and time. Recently, geo-tagged photos from social media have been used as reference data. This procedure has lower costs, but the process of interpreting geo-tagged photos is still time-consuming. In fact, social media point of interest (POI) data, including geo-tagged photos, may contain useful textual information for land cover validation. However, this kind of special textual data has seldom been analysed or used to support land cover validation. This paper examines the potential of textual information from social media POIs as a new reference source to assist in artificial surface validation without photo recognition and proposes a validation framework using modified decision trees. First, POI datasets are classified semantically to divide POIs into the standard taxonomy of land cover maps. Then, a decision tree model is built and trained to classify POIs automatically. To eliminate the effects of spatial heterogeneity on POI classification, the shortest distances between each POI and both roads and villages serve as two factors in the modified decision tree model. Finally, a data transformation based on a majority vote algorithm is then performed to convert the classified points into raster form for the purposes of applying confusion matrix methods to the land cover map. Using Beijing as a study area, social media POIs from Sina Weibo were collected to validate artificial surfaces in GlobeLand30 in 2010. A classification accuracy of 80.68% was achieved through our modified decision tree method. Compared with a classification method without spatial heterogeneity, the accuracy is 10% greater. This result indicates that our modified decision tree method displays considerable skill in classifying POIs with high spatial heterogeneity. In addition, a high validation accuracy of 92.76% was achieved, which is relatively close to the official result of 86.7%. These preliminary results indicate that social media POI datasets are valuable ancillary data for land cover validation, and our proposed validation framework provides opportunities for land cover validation with low costs in terms of money and time.     

Validation and comparison of 1 km global land cover products in China

Four 1 km global land cover products are currently available to the scientific community: the University of Maryland (UMD) global land cover product, the International Geosphere–Biosphere Programme Data and Information System Cover (IGBP‐DISCover), the MODerate resolution Imaging Spectrometer (MODIS) global land cover product and Global Land Cover 2000 (GLC2000). Because of differences in data sources, temporal scales, classification systems and methodologies, it is important to compare and validate these global maps before using them for a variety of studies at regional to global scales. This study aimed to perform the validation and comparison of the four global land cover datasets, and to examine the suitability and accuracy of different coarse spatial resolution datasets in mapping and monitoring cropland across China. To meet this objective, we compared the four global land cover products with the National Land Cover Dataset 2000 (NLCD‐2000) at three scales to evaluate the accuracy of estimates of aggregated cropland areas in China. This was followed by a spatial comparison to assess the accuracies of the four products in estimating the spatial distribution of cropland across China. A comparative analysis showed that there are varying levels of apparent discrepancies in estimating the cropland of China between these four global land cover datasets, and that both area totals and spatial (dis)agreement between them vary from region to region. Among these, the MODIS dataset has the best fit in depicting China's croplands. The coarse spatial resolution and the per pixel classification approach, as well as landscape heterogeneity, are the main reasons for the large discrepancies between the global land cover datasets tested and the reference data.     

Settlement detection and impervious surface estimation in the Mekong Delta using optical and SAR remote sensing data

Information about the extent of impervious surface and its rate of development is a valuable indicator of urban growth and environmental quality and thus relevant for a wide range of research related to urban ecosystems. Using SPOT-5 data from 2005 to 2009, impervious surface was estimated at a subpixel level for the area of Can Tho province in the Mekong Delta, based on a Support Vector Regression model. Training data comprised a set of SPOT-5 reflectance values each associated with an individual value of subpixel imperviousness as their respective labels. The latter were obtained on the basis of a land cover map, which in turn was derived from a pansharpened QB subset by means of an object-oriented image classification approach. In addition, by varying different sets of training data in the model building process the spectral interrelationships between the urban land cover classes (water, bare soil, vegetation, and impervious surface) and their effect on the estimation of subpixel imperviousness could be examined. In order to exclude irrelevant areas (natural/undeveloped land) from the impervious surface estimation process, single-polarised TerraSAR-X data were used to delineate settlement areas by an object-oriented image classification approach. Furthermore, a change detection method was applied for the respective time period in order to test the suitability of the approach for the automated detection of structural developments within the urban topography. Settlement areas were correctly identified with overall accuracies between 81% and 94%, whilst the comparison of the modelled impervious estimates to the training values gave an absolute mean error below 15%. The results prove the suitability of the approach for an area-wide but selective mapping and monitoring of impervious surface cover within settlement areas only.     

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.     

The effect of seasonal variation on automated land cover mapping from multispectral airborne laser scanning data

Multispectral airborne laser scanning (MS-ALS) sensors are a new promising source of data for automated mapping methods. Finding an optimal time for data acquisition is important in all mapping applications based on remotely sensed datasets. In this study, three MS-ALS datasets acquired at different times of the growing season were compared for automated land cover mapping and road detection in a suburban area. In addition, changes in the intensity were studied. An object-based random forest classification was carried out using reference points. The overall accuracy of the land cover classification was 93.9% (May dataset), 96.4% (June) and 95.9% (August). The use of the May dataset acquired under leafless conditions resulted in more complete roads than the other datasets acquired when trees were in leaf. It was concluded that all datasets used in the study are applicable for suburban land cover mapping, however small differences in accuracies between land cover classes exist.     

基于Landsat 8卫星影像的北京地区土地覆盖分类

目的 土地覆盖分类能为生态系统模型、水资源模型和气候模型等提供重要信息,遥感技术运用于土地覆盖分类具有诸多优势。作为区域性土地覆盖分类应用的重要数据源,Landsat 5/7的TM和ETM+等数据已逐渐失效,Landsat 8陆地成像仪(OLI)较TM和ETM+增加了新的特性,利用Landsat 8数据进行北京地区土地覆盖分类研究,探讨处理方法的适用性。方法 首先,确定研究区域内土地覆盖分类系统,并对Landsat 8多光谱数据进行预处理,包括大气校正、地形校正、影像拼接及裁剪;然后,利用灰度共生矩阵提取全色波段纹理信息,与多光谱数据进行融合;最后,使用支持向量机(SVM)进行分类,获得土地覆盖分类结果。结果 经过精度评价和分析发现,6S模型大气校正和C模型地形校正预处理提高了不同类别之间的可分性,多光谱数据结合全色波段纹理特征能有效提高部分地物的土地覆盖分类精度,总体精度提高2.8%。结论 相对于Landsat TM/ETM+数据,Landsat 8 OLI数据新增特性有利于土地覆盖分类精度的提高。本文方法适用于Landsat 8 OLI数据土地覆盖分类研究与应用,能够满足大区域土地覆盖分类应用需求。