利用多平台遥感数据复合研究城市扩展的动态监测

本文利用不同时期的多平台遥感数据,研究了城市动态扩展的监测方法。文中采用了分类结果复合方法,在高速发展的岳阳地区的试验中,取得了良好的效果。在遥感图像的城市信息提取方面,设计了多个不同大小的权系数模板,提取相应的频数矢量作为城市的结构信息。利用这种结构信息,对城市图像进行逐步的结构处理,最终精度在91.7％以上。试验表明,本方法不仅对于一个实用的动态监测系统是一种灵活、实用性较强的监测方法,而且是解决充分利用已有遥感信息资源的有效手段。     

迁移学习支持下的遥感影像对象级分类样本自动选择方法

面向遥感大范围应用的目标,自动化程度仍是遥感影像分类面临的重要问题,样本的人工选择难以适应当前土地覆盖信息自动化提取的实际应用需求。为了构建一套基于先验知识的遥感影像全自动分类流程,本文将空间信息挖掘技术引入到遥感信息提取过程中,提出了一种面向遥感影像对象级分类的样本自动选择方法。该方法通过变化检测将不变地物标示在新的目标影像上,并将过去解译的地物类别知识迁移至新的影像上,建立新的特征与地物关系,从而完成历史专题数据辅助下目标影像的自动化的对象级分类。实验结果表明,在已有历史专题层的图斑知识指导下,该方法能有效地自动选择适用于新影像分类的可靠样本,获得较好的信息提取效果,提高了对象级分类的效率。     

基于自适应共振模型的遥感影像分类方法研究

人工神经网络（ANN）是人视觉和服的基本功能的抽象、简化和模拟。在对遥感影像的综合解释应用中，与传统的统计方法和符号逻辑方法相比较，ANN更接近人对影像的视觉解译分析过程。自适应共振理论（ART）是一种自组织产生认知编码的神经网络理论，其自组织、反馈式增量学习机能，能兼顾适应性和稳定性，克服了一般神经网络学习速度慢、网络结构难以确定、局部最小陷阱等缺陷。以FUZZY－ART和ARTMAP为基础，提出基于ART遥感影像非监督和监督分类的一般模型，并以实际上土覆盖分类和城市结构信息提取为应用实例，通过与传统统计方法和一般ANN分类器相比较，ART具有正确率更同、学习速度快、自适应性等优点，是复杂数据分类和信息提取的有效工具。     

Land Use/Cover Change Detection of Tiruchirapalli City, India, Using Integrated Remote Sensing and GIS Tools

The rapid growth of urban population in India is a cause of concern among country??s urban and town planners for efficient urban planning. The drastic growth of urban areas has resulted in sharp land use and land cover changes. In recent years, the significance of spatial data technologies, especially the application of remotely sensed data and geographical information systems (GIS) has been widely used. The present study investigates the urban growth of Tiruchirapalli city, Tamilnadu using IRS satellite data for the years 1989, 1992, 1995, 1998, 2001, 2004, 2007, and 2010. The eight satellite images are enhanced using convolution spatial enhancement method with Kernel (7?×?7) edge enhance function. Supervised classification method is used to classify the urban land use and land cover. The GIS is used to prepare the different layers belonging to various land uses identified from remotely sensed data. The analysis of the results show the drastic increase of built up area and reduced green cover within the city boundary limit.     

GIS支持下珠江三角洲城镇建筑覆盖变化遥感监测分析

从珠江三角洲区域总体出发，探讨大范围城镇建筑覆盖变化的遥感监测方法。分析了城镇建筑覆盖变化专题信息提取方法概况，比较了最大似然法、上下文分类法、人工神经网络三种分类方法的优缺点。借助GIS技术的术持，在计算机自动分类的基础上利用专家知识改进了分类结果。根据遥感监测结果，对1988—1998年间珠江三角洲城镇建筑覆盖变化特征进行了分析。     

Development of a remotely sensed,historical land-cover change database for rural Chihuahua,Mexico

The purpose of this article is to describe the development of a remotely sensed, historical land-cover change database for the northwestern quarter of Chihuahua, Mexico, The database consists of multi-temporal land-cover classifications and change detection images. The database is developed to facilitate future investigations that examine urban–rural linkages as possible drivers of rural land-use and land-cover changes. To develop the needed land-cover change database, this study uses the North American Landsat Characterization (NALC) MSS triplicates because of their temporal depth and spatial breadth. Challenges exist, however, to effective classification and change detection using the NALC triplicates, including illumination differences across multiple scenes and periods caused by topographic and solar variations and the lack of ground reference data for historic periods. Therefore, creation of the database is a four step process. First, extensive pre-processing is performed to enhance comparability of multi-date images. Pre-processing includes topographic correction, mosaic creation and multi-date radiance normalization. Second, ancillary sources of land-cover data are combined with visual interpretations of enhanced images to define reference pixels used to classify the images using the maximum likelihood algorithm. Third, classification accuracy is assessed. Fourth, post-classification change detection is performed. Results indicate significant image improvements after pre-processing that permit very good overall classification (86.26% classified correctly) and change detection. To conclude, findings are presented that indicate significant changes to arid grasslands/shrublands and forest resources in mountainous regions.     

Crowdsourcing functions of the living city from Twitter and Foursquare data

ABSTRACT

Urban functions are closely related to people’s spatiotemporal activity patterns, transportation needs, and a city’s business distribution and development trends. Studies investigating urban functions have used different data sources, such as remotely sensed imageries, observation, photography, and cognitive maps. However, these data sources usually suffer from low spatial, temporal, and thematic resolution. This article attempts to investigate human activities to understand urban functions through crowdsourcing social media data. In this study, we mined Twitter and Foursquare data to extract and analyze six types of human activities. The spatiotemporal analysis revealed hotspots for different activity intensities at different temporal resolution. We also applied the classified model in a real-time system to extract information of various urban functions. This study demonstrates the significance and usefulness of social sensing in analyzing urban functions. By combining different platforms of social media data and analyzing people’s geo-tagged city experience, this article contributes to leverage voluntary local knowledge to better depict human dynamics, discover spatiotemporal city characteristics, and convey information about cities.     

高分辨率多光谱影像城区建筑物提取研究

城区高空间分辨率遥感数据由于存在大量同物异谱和异物同谱现象,应用传统的基于像元光谱分类的方法进行建筑物分类提取难以取得满意的效果。本文发展了一种从高分辨率Ikonos卫星影像上基于知识规则的面向对象分类提取城区建筑物方法,包括如下步骤：(1)融合1m全色和4m多光谱波段影像,生成1m分辨率的多光谱融合影像;（2）分割融合影像;（3）执行基于对象光谱的最近邻监督分类;（4）应用模糊逻辑分类器结合光谱、空间、纹理和上下文特征等知识规则进行建筑物分类。精度统计结果表明,本文提出的分类方法提取城区建筑物取得了93%的精度。     

遥感信息图谱支持的土地覆盖自动分类

遥感数据的海量堆积与应用信息的匮乏日益凸显信息认知提取的重要性,在地学信息图谱方法论的指导下,同时参考视觉认知流程,提出了遥感信息图谱认知方法用于遥感数据的自动解译。在地理信息系统的统一框架下逐步挖掘多源遥感数据的"图"、"谱"特征并进行图谱耦合分析,通过多尺度分割、特征分析、监督学习等关键步骤完成"察觉—分辨—确认"的地学认知流程,初步满足自动化和智能化应用需求。在土地覆盖信息自动解译应用中建立了基于"图谱"先验知识的管理与运用机制以实现自动化,采用机器学习算法提升智能化程度,并以自适应迭代控制模型使结果精度向最优逼近。选取了珠江三角洲的试验区域进行了基于ALOS多光谱影像的土地覆盖自动分类,结果符合预期,说明了本文方法的可行性。     

遥感大数据智能解译的地理学认知模型与方法

随着遥感数据和计算机算力的爆炸式增长、智能分析算法瓶颈的突破,亟须提升与之相匹配的遥感大数据处理与分析能力。针对复杂场景下遥感大数据智能处理与地理学认知耦合关联和交叉融合的关键问题,本文分析了遥感大数据与地理科学各自的特点与相互关系,提出了多模态知识融合关联的深度网络构建和面向地理制图的遥感智能解译思路,建立了遥感大数据智能处理与应用体系框架;面向技术发展和行业应用,本文提出了分别建设通用高分辨率遥感智能处理系统和智能精准应用平台的总体路线,以期推动遥感智能解译技术创新和工程化应用的全面发展。     

Examining the impacts of urbanization on surface radiation using Landsat imagery

Metropolitan Beijing is facing many environmental problems such as haze and urban heat island due to the rapid urbanization. Surface shortwave, longwave, and net radiations are key components of the surface-atmosphere radiation budget. Since megacities are affected by the thermal radiation of complex landscape structures and atmospheric environments, quantitative and spatially explicit retrieval from remotely sensed data remains a challenge. We collected the surface radiation fluxes from seven fixed sites representing different land-use types to calibrate the local parameters for remotely sensed retrieval of net radiation. We proposed a remote sensing–based surface radiation retrieval method by embedding the underlying land covers and integrating the observational data. The improved method is feasible to accurately retrieve surface radiation and delineate spatial characteristics in metropolitan areas. The accuracy evaluation indicated that the difference between remotely sensed and in situ observed net radiation ranged within 0~± 40 W· m^?2. The root mean squared error of the estimated net surface radiation was 32.71 W· m^?2. The strongly spatial heterogeneity of surface radiation components in metropolitan Beijing was closely related to land-cover patterns from urban area to outskirts. We also found that the surface net radiation had a decreasing trend from 1984 to 2014, and the net radiation in the urban area was lower than that in the outskirts. According to the surface radiation budgets, urbanization resulted in the cooling effect in net radiation flux in the daytime, which was stemmed from low atmospheric transmittances from massive aerosol concentration and high surface albedo from light building materials.     

遥感信息处理不确定性的可视化表达

如何全面、准确地度量和可视化表达遥感信息处理中不确定性的程度和空间分布方式,是遥感信息不确定性研究的关键问题之一.传统的度量方法(例如误差矩阵)是将以训练样本集为基础的度量作为总分类精度的度量,而我们需要估计模型对于"样本外数据"的性能.本文首先利用信息论和粗糙集理论等度量遥感分类影像属性信息的不确定性,提出基于像元、目标和影像的遥感信息不确定性度量指标;然后分别描述了基于不同度量指标的可视化表达方式,并对我国黄河三角洲地区的Landsat TM影像进行了分类信息不确定性度量和可视化表达实验.     

城市用地空间信息分类与数字重建——以长春百年城市内部用地变化为例

针对单一应用遥感影像难以进行城市内部用地结构分类以及高精度城市内部用地多期空间数据叠置分析中位置误差问题建立了基于"分层分类"与"对象分割"的城市内部用地空间信息数字重建方法。实现对特大城市产业用地(住宅、商业、工业等)以及交通、水系、生态绿地等不同功能结构用地的高精度监测以及历史演变过程的重建。综合集成SPOT5,1︰1万地形图、历史地图及城市规划图等辅助信息对长春城市1905年以来城市用地信息进行分类。研究表明,在专家知识参与下人—机交互解译,集成多源空间信息对实现高精度城市用地空间信息重建具有较高的应用价值,该方法不仅能提高城市用地分类精度而且能提高城市用地空间信息提取效率以及多期空间数据叠置分析的定位精度。     

Change detection assessment using fuzzy sets and remotely sensed data: an application of topographic map revision

This paper explores a methodology for computing the amount of changes that have occurred within an area by using remotely sensed technologies and fuzzy modelling. The discussion concentrates on the formulation of a standard procedure that, using the concept of fuzzy sets and fuzzy logic, can define the likelihood of changes detected from remotely sensed data. Furthermore, an example of how fuzzy visualisation of areas undergoing changes can be incorporated into a decision support system for prioritisation of areas requiring topographic map revision and updating is presented. By adapting the membership function of the fuzzy model to fit the shape of the histogram characterising the change image (derived from any of the common pre-classification methods of change detection), areas can be identified according to their likelihood of having undergone change during the period of observation.     

Remote sensing research issues of the National Land Use Change Program of China

In 1999, the Ministry of Land and Resources (MLR) of China launched the National Land Use Change Program especially to monitor the scale and distribution of urban expansion and the decrease in cultivated land through remote sensing technology. This Program has been carried out annually and continuously for seven years since then and played an important role in the policy-making of MLR about land management and planning. This paper gives an overview about this Program and discusses several research issues. First, the remote sensing data sources and other ancillary data used in this Program are presented. The approaches for image preprocessing, i.e. radiometric normalization, image geometric rectification and image fusion are then introduced with an emphasis on the algorithm development for image registration. Second, land use change detection technique is the most critical and complex aspect of the Program. The methodologies for change detection using either bi-temporal image pair or one existing land use map and one remotely sensed image are detailed. Third, since the data of land use changes derived from remote sensing will be operationally used for local and central government, field validation and accuracy assessment are crucial to ensure the reliability of change detection results. The strategy of field work and the resulting accuracy evaluations is presented. The land use and change information derived from remotely sensed data has wide applications for land management, including land use database updating, verification of land use planning and monitoring of national high-tech parks. Last, suggestions on how to make full use of the images and change detection result, to improve the consistency of land use classification and to develop change detection algorithms for diverse and complex remote sensing data are given.     

Quantifying annual land-cover change and vegetation greenness variation in a coastal ecosystem using dense time-series Landsat data

Land-cover change may affect water and carbon cycles when transitioning from one land-cover category to another (land-cover conversion, LCC) or when the characteristics of the land-cover type are altered without changing its overall category (land-cover modification, LCM). Given the increasing availability of time-series remotely sensed data for earth monitoring, there has been increased recognition of the importance of accounting for both LCC and LCM to study annual land-cover changes. In this study, we integrated 1,513 time-series Landsat images and a change-updating method to identify annual LCC and LCM during 1986–2015 in the coastal area of Zhejiang Province, China. The purpose was to quantify their contributions to land-cover changes and impacts on the amount of vegetation. The results show that LCC and LCM can be successfully distinguished with an overall accuracy of 90.0%. LCM accounted for 22% and 40.5% of the detected land-cover changes in reclaimed and inland areas, respectively, during 1986–2015. In the reclaimed area, LCC occurred mostly in muddy tidal flats, construction land, aquaculture ponds, and freshwater herbaceous land, whereas LCM occurred mostly in freshwater herbaceous land, Spartina alterniflora, and muddy tidal flats. In the inland area, both LCC and LCM were concentrated in forest and dryland. Overall, LCC had a mean magnitude of normalized difference vegetation index (NDVI) change similar to that of LCM. However, LCC had a positive effect and LCM had a negative effect on NDVI change in the reclaimed area. Both LCC and LCM in the inland area had negative impacts on vegetation greenness, but LCC resulted in larger NDVI change magnitude. Impacts of LCC and LCM on vegetation greenness were quantified for each land-cover type. This study provided a methodological framework to take both LCC and LCM into account when analyzing land-cover changes and quantified their effects on coastal ecosystem vegetation.     

Change detection from remotely sensed images: From pixel-based to object-based approaches

The appetite for up-to-date information about earth’s surface is ever increasing, as such information provides a base for a large number of applications, including local, regional and global resources monitoring, land-cover and land-use change monitoring, and environmental studies. The data from remote sensing satellites provide opportunities to acquire information about land at varying resolutions and has been widely used for change detection studies. A large number of change detection methodologies and techniques, utilizing remotely sensed data, have been developed, and newer techniques are still emerging. This paper begins with a discussion of the traditionally pixel-based and (mostly) statistics-oriented change detection techniques which focus mainly on the spectral values and mostly ignore the spatial context. This is succeeded by a review of object-based change detection techniques. Finally there is a brief discussion of spatial data mining techniques in image processing and change detection from remote sensing data. The merits and issues of different techniques are compared. The importance of the exponential increase in the image data volume and multiple sensors and associated challenges on the development of change detection techniques are highlighted. With the wide use of very-high-resolution (VHR) remotely sensed images, object-based methods and data mining techniques may have more potential in change detection.     

基于频域滤波的高分辨率遥感图像城市河道信息提取

提出一种基于频域滤波的城市河道信息提取方法。首先对高分辨率遥感图像进行傅里叶变换得到频谱图, 并利用径向和角向分布图分析城市河道的频谱特征。其次, 基于城市河道的双线型特点, 将其分为边缘特征和低频信息两个部分, 并根据周期性纹理的频谱模型和地物频谱能量分布规律确定两个部分的频域识别标志。然后设计相应的扇环形带通log Butterworth滤波器和低通Butterworth滤波器分别对城市河道的边缘特征和低频信息进行提取, 并根据该两部分信息实现城市河道信息提取。最后对城市河道信息提取结果进行定量评价, 结果表明, 本文方法可以有效地实现城市河道的信息提取。     

Semantic segmentation of high spatial resolution images with deep neural networks

Availability of reliable delineation of urban lands is fundamental to applications such as infrastructure management and urban planning. An accurate semantic segmentation approach can assign each pixel of remotely sensed imagery a reliable ground object class. In this paper, we propose an end-to-end deep learning architecture to perform the pixel-level understanding of high spatial resolution remote sensing images. Both local and global contextual information are considered. The local contexts are learned by the deep residual net, and the multi-scale global contexts are extracted by a pyramid pooling module. These contextual features are concatenated to predict labels for each pixel. In addition, multiple additional losses are proposed to enhance our deep learning network to optimize multi-level features from different resolution images simultaneously. Two public datasets, including Vaihingen and Potsdam datasets, are used to assess the performance of the proposed deep neural network. Comparison with the results from the published state-of-the-art algorithms demonstrates the effectiveness of our approach.     

决策级融合的多分辨率遥感影像变化检测

针对多时相、多分辨率遥感影像数据的特点,充分考虑不同分辨率数据和不同变化检测应用的需求,将由粗到精数据集分层检测和决策级融合的思想引入到变化检测,以多时相多分辨率ALOS遥感影像为例,构建并试验了由粗到精变化检测的技术流程.该方法将ALOS多光谱数据视为粗数据集,将全色数据和融合数据视为精数据集,通过对3种数据集变化检...