遥感影像与POI数据相结合的城市建成区提取适用性研究——以沈阳市为例

城市建成区的发展状况是地理国情监测的重要内容，本文基于遥感影像数据和POI数据对城市建成区进行提取，针对二者的适用性问题进行了研究。试验以沈阳市为研究区域，在研究区域内选择2016年遥感影像数据和POI数据作为数据源进行对比分析。首先，对遥感影像数据和POI数据进行预处理；其次，通过监督分类的方法对遥感影像进行建成区的提取；然后，采用核密度估计法分析POI数据并提取出建成区；最后，利用叠加分析法对比分析这两种数据的适用性。试验结果表明：使用遥感影像数据作为数据源可以较为全面客观地反映城市建成区的发展现状；利用POI数据提取出的城市建成区具有较强的经济属性，能够很好地反映出城市中的经济活跃区。     

光谱特征扩展的时间序列Landsat数据地表覆盖分类

时间序列遥感影像常用于地表覆盖监测及其变化监测。然而,利用时序遥感数据—尤其是中分辨率遥感数据监测地表覆盖变化,其方法基本是先对多期影像分别进行监督分类然后对比分类结果。由于这种方法需要对每期遥感影像单独选择分类训练样本,而对于历史影像,常常难以获得可靠的样本数据。本文基于遥感数据定量化处理,尝试利用光谱特征扩展方法对时间序列Landsat数据进行分类:首先,结合一种新的大气校正方法和相对辐射归一化方法,对时间序列Landsat数据进行定量化处理,以消除各期影像之间的辐射差异,获得地表反射率数据。然后,论文选择一期易于获得分类训练样本的反射率数据作为"参考影像",并结合样本数据提取不同地表覆盖类型的光谱特征。最后,将"参考影像"中提取的地物光谱特征,扩展到所有时间序列反射率数据进行分类。论文利用青藏高原玛多地区的5景Landsat数据对本文的方法进行了验证,结果显示:基于光谱特征扩展的分类方法,可有效对定量化处理后的Landsat数据进行分类,分类总体精度为88.35%—94.25%,分类结果和传统的单景监督分类结果具有较好的一致性。此外,研究也发现,"参考影像"和待分类图像获取时间的季相差异会影响其分类的精度。     

基于决策树分类的地表覆盖遥感信息提取

地处西南的渝北地区地表覆盖类型复杂、土地利用多元化,仅依赖于光谱特征的传统遥感信息提取方法难以获得较高的分类精度。利用决策树分类技术对渝北地区的TM遥感影像进行分类,除光谱信息外还结合地质、NDVI、PCI等多源数据进行实验。结果表明,总精度和Kappa系数分别为88.42%和0.854 7,较传统的监督分类和仅依赖于光谱特征的决策树分类方法有较大提高,这也表明基于多源数据的决策树分类技术对地表覆盖复杂地区的遥感影像分类比较适用,是遥感信息提取的一种有效手段。     

面向遥感影像智能分类的海量样本数据采集方法

以地理国情监测高分辨率遥感影像及高精度地表覆盖分类产品为数据源，提出了一种面向遥感影像智能分类、基于位置匹配技术的全国尺度海量样本数据采集方法。根据数据源特征，研究了县域采集数量权重设置、坐标投影转换、栅格灰度重采样、无效样本数据过滤、地表覆盖分类码映射、样本数据命名标识、特定地表覆盖类型样本数据采集等关键技术，构建了位置匹配的遥感影像数据与分类标签数据组成的样本数据对，开发了样本数据自动采集软件。利用该方法，以县级行政区划为单元，实现了全国尺度海量样本数据采集。选取其中5个县域的成果，评估了方法的实用性及运算性能。研究表明：该方法提升了生产全国尺度海量样本数据的计算响应速度；采集的样本数据能够满足遥感影像智能分类对样本源高质量、大规模的需求，提升了遥感影像分类与预测的准确度。     

多尺度特征融合神经网络的地表覆盖提取

深度学习的快速发展,为高分辨率卫星遥感影像解译提供了更好的技术手段和应用前景。围绕高分卫星遥感影像地表覆盖信息提取,利用高分辨率卫星遥感影像制作5种常见的地表覆盖类型的像素级样本数据集,并提出一种基于注意力增强与多尺度特征融合的语义分割方法,实现地表覆盖自动提取。通过影像波段选择、预训练模型迁移学习、损失函数改进等方法,提升语义分割模型识别精度,最优的提取结果中,5种地表覆盖类型的F₁均值、IoU均值和总体精度分别达到了78.6%、66.8%、85.0%,除道路之外,耕草、林地、建筑、水体的F₁均超过80%,且分类图斑边界能够与影像中的地物边界很好套合。实验表明建立的卫星影像地表覆盖分类样本数据集和分类方法,能够应用于高分辨率卫星影像地表覆盖信息提取。     

高分辨率遥感影像下沿海地区地表覆盖信息的提取

沿海地区地表覆盖信息是全国地理国情普查的重要内容,遥感影像分类技术为沿海地区地表覆盖信息提供了一种重要方法。本文基于GF-1高分辨率遥感影像,建立了沿海地区地表覆盖分类系统,采用中国测绘科学研究院自主研发的面向对象GLC决策树分类方法和软件进行了地表覆盖分类。通过对某试验区进行分类试验,并结合该区地表覆盖标准分类图进行精度评价,验证了基于高分辨率影像,面向对象GLC决策树分类方法在沿海地区地表覆盖信息提取上的有效性及优越性,其总体分类精度和Kappa系数分别为87.201 8%、0.840 6,均高于SVM分类法。最后提出基于高分辨率遥感影像的沿海地区地表覆盖信息提取流程。     

大比例尺地形图在地理国情普查中的应用探讨

地理国情普查最主要的内容是地表覆盖分类与重要地理国情要素,其中地表覆盖主要基于遥感或航空影像影像进行分割、分类,解译出地表覆盖物的类型,以便从地理的角度进行统计分析。大比例尺地形图具有地表覆盖物与地理要素丰富信息的特点,本文结合生产实践探索出一套大比例尺地形图数据应用于地理国情普查的有效技术方法。     

人工神经网络遥感影像分类模型及其与知识集成方法研究

以多层感知器（MLP）为例，探讨了地学知识与ANN融合进行遥感影像分类的方法。首先对MLP网络结构、学习算法及其改进进行分析；然后总结了MLP进行遥感影像分类的一般方法和存在的缺陷；发展了基于知识的MLP神经网络遥感影像分类模型，并具体利用基于规则的MLP方法进行了遥感土地覆盖分类的实验，把获得的结果与传统统计方法与一般ANN方法进行了综合比较，获得了有意义的结果。     

基于CBERS-02遥感影像的湿地地表覆被分类研究

基于CBERS遥感影像多光谱数据,运用信息量、相关系数及OIF方法,分析了波段数据特征,获得了对影像数据的整体认识。 通过绘制地表覆被类型的样本均值光谱曲线,分析了多种地表覆被类型在5个影像波段中的光谱特征,得出了不同覆被类型在各波 段中的反射特性。基于典型地表覆被类型样本数据,分别针对影像的5个波段及第一主成分、归一化植被指数这两个重要特征,运 用盒须图进一步分析了不同地表覆被类型的分异特性。运用Z-test统计方法,筛选出了区分不同地表覆被类型的最优纹理特征。运 用面向对象分类技术开展了研究区覆被类型分类实验,验证了基于CBERS遥感影像进行内陆淡水湿地区地表覆被分类的可行性,丰 富了地表覆被信息提取方法,拓展了CBERS遥感影像的应用领域。     

基于面向对象与规则的Sentinel-2A影像土地覆被分类——以江西省都昌县为例

土地覆被作为地表自然和人工建造物的综合体，是开展土地科学相关研究的重要基础，在遥感大数据背景下，准确、快速、自动化进行土地覆被提取技术一直是遥感研究中的重点。本文基于eCognition软件，采用面向对象的多尺度分割法，综合考虑地物在遥感影像上的光谱、形状和纹理特征，建立多种地物提取规则。通过模糊函数、支持向量机（SVM）和阈值法对研究区的土地覆被进行分类提取，并与研究区的FROM-GLC10数据和土地利用变更数据进行了对比分析。结果表明：①研究区土地覆被分类的总体精度为97%，Kappa系数为0.96，分类精度较高；②基于10 m分辨率影像，综合使用形状、纹理、光谱信息对于道路的提取具有较好的效果，道路提取Kappa系数为0.84；③分类结果在面积和空间分布上都优于FROM-GLC10数据，与研究区实际土地变更数据保持较好的一致性。基于面向对象与规则的分类方法提取地物能够有效利用多种遥感影像特征，分类精度高，对于处理高分辨率遥感数据具有很好的优势。     

基于多尺度影像分割的面向对象城市土地覆被分类研究——以马来西亚吉隆坡市城市中心区为例

城市受人类活动影响比较大,结构组成比较复杂,对该区域进行分类研究存在一些问题。甚高分辨率遥感影像,以其丰富的细节信息为城市土地覆被分类研究提供了可能。本文结合使用甚高分辨率QuickBird遥感影像和激光扫描LIDAR数据,论述了利用多尺度、多变量影像分割的面向对象的分类技术对马来西亚基隆坡市城市中心区的土地覆被分类研究。针对特定地物选择合适的影像分割特征和分割尺度、按照合理的提取顺序逐步进行城市土地覆被信息提取。在建筑物的提取过程中构建了归一化数字表面模型nDSM,使用成员函数将建筑物信息提取出来。精度评价结果表明,利用该方法得到了理想的城市土地覆被分类结果,其分类总精度从常规面向对象分类方法的83.04%上升到88.52%,其中建筑物生产精度从60.27%增加到93.91%。     

Extracting urban functional regions from points of interest and human activities on location‐based social networks

Data about points of interest (POI) have been widely used in studying urban land use types and for sensing human behavior. However, it is difficult to quantify the correct mix or the spatial relations among different POI types indicative of specific urban functions. In this research, we develop a statistical framework to help discover semantically meaningful topics and functional regions based on the co‐occurrence patterns of POI types. The framework applies the latent Dirichlet allocation (LDA) topic modeling technique and incorporates user check‐in activities on location‐based social networks. Using a large corpus of about 100,000 Foursquare venues and user check‐in behavior in the 10 most populated urban areas of the US, we demonstrate the effectiveness of our proposed methodology by identifying distinctive types of latent topics and, further, by extracting urban functional regions using K‐means clustering and Delaunay triangulation spatial constraints clustering. We show that a region can support multiple functions but with different probabilities, while the same type of functional region can span multiple geographically non‐adjacent locations. Since each region can be modeled as a vector consisting of multinomial topic distributions, similar regions with regard to their thematic topic signatures can be identified. Compared with remote sensing images which mainly uncover the physical landscape of urban environments, our popularity‐based POI topic modeling approach can be seen as a complementary social sensing view on urban space based on human activities.     

An Automated Approach for Land Cover Classification Based on a Fuzzy Supervised Learning Framework

This paper proposes an automatic framework for land cover classification. In majority of published work by various researchers so far, most of the methods need manually mark the label of land cover types. In the proposed framework, all the information, like land cover types and their features, is defined as prior knowledge achieved from land use maps, topographic data, texture data, vegetation’s growth cycle and field data. The land cover classification is treated as an automatically supervised learning procedure, which can be divided into automatic sample selection and fuzzy supervised classification. Once a series of features were extracted from multi-source datasets, spectral matching method is used to determine the degrees of membership of auto-selected pixels, which indicates the probability of the pixel to be distinguished as a specific land cover type. In order to make full use of this probability, a fuzzy support vector machine (SVM) classification method is used to handle samples with membership degrees. This method is applied to Landsat Thematic Mapper (TM) data of two areas located in Northern China. The automatic classification results are compared with visual interpretation. Experimental results show that the proposed method classifies the remote sensing data with a competitive and stable accuracy, and demonstrate that an objective land cover classification result is achievable by combining several advanced machine learning methods.     

A multi‐modal transportation data‐driven approach to identify urban functional zones: An exploration based on Hangzhou City,China

Recent urban studies have used human mobility data such as taxi trajectories and smartcard data as a complementary way to identify the social functions of land use. However, little work has been conducted to reveal how multi‐modal transportation data impact on this identification process. In our study, we propose a data‐driven approach that addresses the relationships between travel behavior and urban structure: first, multi‐modal transportation data are aggregated to extract explicit statistical features; then, topic modeling methods are applied to transform these explicit statistical features into latent semantic features; and finally, a classification method is used to identify functional zones with similar latent topic distributions. Two 10‐day‐long “big” datasets from the 2,370 bicycle stations of the public bicycle‐sharing system, and up to 9,992 taxi cabs within the core urban area of Hangzhou City, China, as well as point‐of‐interest data are tested to reveal the extent to which different travel modes contribute to the detection and understanding of urban land functions. Our results show that: (1) using latent semantic features delineated from the topic modeling process as the classification input outperforms approaches using explicit statistical features; (2) combining multi‐modal data visibly improves the accuracy and consistency of the identified functional zones; and (3) the proposed data‐driven approach is also capable of identifying mixed land use in the urban space. This work presents a novel attempt to uncover the hidden linkages between urban transportation patterns with urban land use and its functions.     

基于MODIS数据的环北京地区土地资源监测研究

本文基于MODIS 16天合成的NDVI时间序列数据及其他辅助数据,首先用PCA方法对NDVI时间序列数据进行信息增强与压缩处理,结合LST数据、DEM数据及降雨温度数据,利用模糊K-均值非监督分类法,进行环北京地区的土地覆盖分类,得到土地资源现状情况。然后利用变化矢量(CVA)分析方法对环北京地区的土地利用及植被覆盖的多年变化状况进行了分析。结果表明,MODIS数据能很好的应用于大范围的土地资源监测中,并能得到较好的结果。     

土地覆盖制图：基于最优化遥感数据的支撑向量机分类

遥感数据具有在不同空间、光谱和时间尺度上获取地表测量信息的能力,使其成为获取土地覆盖信息的一个主要数据源。影像分类即把卫星影像上的相关像元划分给某类已知的土地覆盖类型的过程。支撑向量机（SVMs）是一种土地覆盖分类的新技术。三种常用的SVMs是：基于线性和多项式的SVM以及具有高斯核函数的SVM分类器,分类能否成功地应用有赖于其各自选择的最佳参数。但是海量的遥感数据使得这些参数的确定速度十分缓慢。本文研究了一种新的基于最优化遥感数据压缩技术的SVM分类方法。研究显示用于获取SVM参数的数据量能够在不影响土地覆盖的分类精度的前提下进行压缩。数据压缩成功的应用于多项式和高斯核函数的SVM分类,而线性SVM的分类精度却非常低。     

Classifying historical remotely sensed imagery using a tempo-spatial feature evolution (T-SFE) model

Large and growing archives of orbital imagery of the earth’s surface collected over the past 40 years provide an important resource for documenting past and current land cover and environmental changes. However uses of these data are limited by the lack of coincident ground information with which either to establish discrete land cover classes or to assess the accuracy of their identification. Herein is proposed an easy-to-use model, the Tempo-Spatial Feature Evolution (T-SFE) model, designed to improve land cover classification using historical remotely sensed data and ground cover maps obtained at later times. This model intersects (1) a map of spectral classes (S-classes) of an initial time derived from the standard unsupervised ISODATA classifier with (2) a reference map of ground cover types (G-types) of a subsequent time to generate (3) a target map of overlaid patches of S-classes and G-types. This model employs the rules of Count Majority Evaluation, and Subtotal Area Evaluation that are formulated on the basis of spatial feature evolution over time to quantify spatial evolutions between the S-classes and G-types on the target map. This model then applies these quantities to assign G-types to S-classes to classify the historical images. The model is illustrated with the classification of grassland vegetation types for a basin in Inner Mongolia using 1985 Landsat TM data and 2004 vegetation map. The classification accuracy was assessed through two tests: a small set of ground sampling data in 1985, and an extracted vegetation map from the national vegetation cover data (NVCD) over the study area in 1988. Our results show that a 1985 image classification was achieved using this method with an overall accuracy of 80.6%. However, the classification accuracy depends on a proper calibration of several parameters used in the model.     

Mapping the Land Cover of Mexico Using AVHRR Time‐Series Data Sets

Abstract

An important methodological and analytical requirement for analyzing spatial relationships between regional habitats and species distributions in Mexico is the development of standard methods for mapping the country's land cover/land use formations. This necessarily involves the use of global data such as that produced by the Advanced Very High Resolution Radiometer (AVHRR). We created a nine‐band time‐series composite image from AVHRR Normalized Difference Vegetation Index (NDVI) bi‐weekly data. Each band represented the maximum NDVI for a particular month of either 1992 or 1993. We carried out a supervised classification approach, using the latest comprehensive land cover/vegetation map created by the Mexican National Institute of Geography (INEGI) as reference data. Training areas for 26 land cover/vegetation types were selected and digitized on the computer's screen by overlaying the INEGI vector coverage on the NDVI image. To obtain specific spectral responses for each vegetation type, as determined by its characteristic phenology and geographic location, the statistics of the spectral signatures were subjected to a cluster analysis. A total of 104 classes distributed among the 26 land cover types were used to perform the classification. Elevation data were used to direct classification output for pine‐oak and coastal vegetation types. The overall correspondence value of the classification proposed in this paper was 54%; however, for main vegetation formations correspondence values were higher (60‐80%). In order to obtain refinements in the proposed classification we recommend further analysis of the signature statistics and adding topographic data into the classification algorithm.     

Classifiers vs. input variables—The drivers in image classification for land cover mapping

The study investigates the performance of image classifiers for landscape-scale land cover mapping and the relevance of ancillary data for the classification success in order to assess and to quantify the importance of these components in image classification. Specifically tested are the performance of maximum likelihood classification (MLC), artificial neural networks (ANN) and discriminant analysis (DA) based on Landsat7 ETM+ spectral data in combination with topographic measures and NDVI. ANN produced high accuracies of more than 75% also with limited input information, while MLC and DA produced comparable results only by incorporating ancillary data into the classification process. The superiority of ANN classification was less pronounced on the level of the single land cover classes. The use of ancillary data generally increased classification accuracy and showed a similar potential for increasing classification accuracy than the selection of the classifier. Therefore, a stronger focus on the development of appropriate and optimised sets of input variables is suggested. Also the definition and selection of land cover classes has shown to be crucial and not to be simply adaptable from existing land cover class schemes. A stronger research focus towards discriminating land cover classes by their typical spectral, topographic or seasonal properties is therefore suggested to advance image classification.     

Rule‐based Integration of Remotely‐sensed Data and GIS for Land Cover Mapping in NE Costa Rica

Abstract

A classification method was developed for mapping land cover in NE Costa Rica at a regional scale for spatial input to a biogeochemical model (CENTURY). To distinguish heterogeneous cover types, unsupervised classifications of Landsat Thematic Mapper data were combined with ancillary and derived data in an iterative process. Spectral classes corresponding to ground control types were segregated into a storage raster while ambiguous pixels were passed through a set of rules to the next stage of processing. Feature sets were used at each step to help sort spectral classes into land cover classes. The process enabled different feature sets to be used for different types while recognizing that spectral classification alone was not sufficient for separating cover types that were defined by heterogeneity. Spectral data included the TM reflective bands, principal components and the NDVI. Ancillary data included GIS coverages of swamp extents, banana plantation boundaries and river courses. Derived data included neighborhood variety and majority measures that captured texture. The final map depicts 18 land cover types and captures the general patterns found in the region. Some confusion still exists between closely related types such as pasture with different amounts of tree cover.