基于自适应区间二型模糊聚类的遥感土地覆盖自动分类

遥感影像土地覆盖分类面临"类别密度差异显著"、"同谱异物"和"同物异谱"等不确定性问题,传统的分类方法(如FCM)因不能描述高阶模糊不确定性,无法完成准确建模,使分类误差较大,而二型模糊集恰是处理此类不确定性的有效工具.在引入二型模糊集新概念和自适应降型新方法的基础上,提出一种自适应二型模糊分类方法(A-IT2FCM):(1)基于样本集模糊距离度量构建面向分类的区间二型模糊集,以尽可能降低对先验知识和预设参数的依赖,从而满足自动分类的要求;(2)给出一种自适应探求等价一型代表(模糊)集合的高效降型方法,在此基础上进行自适应区间二型模糊聚类.实验数据为珠海横琴和北京颐和园的SPOT5影像数据,对比方法有AIT2FCM、基于Karnik-Mendel算法降型和基于Tizhoosh提出的简易降型方法的区间二型模糊C均值聚类以及作者前期研究提出的区间值模糊C-均值算法(IV-FCM).实验结果表明,A-IT2FCM方法分类效果佳,在类别具有较大密度差异和多重模糊性时能得到比FCM及IV-FCM更精确的边界和更连贯的类别,适于处理遥感影像土地覆盖类别的深层不确定性;同时在"光谱混叠"现象严重时,可以获得比对比方法更稳健、精度更高的影像自动分类结果,且时间复杂度明显低于基于Karnik-Mendel方法.     

Novel fuzzy uncertainty modeling for land cover classification based on clustering analysis

It is well known that there is a degree of fuzzy uncertainty in land cover classification using remote sensing (RS) images. In this article, we propose a novel fuzzy uncertainty modeling algorithm for representing the features of land cover patterns, and present an adaptive interval type-2 fuzzy clustering method. The proposed fuzzy uncertainty modeling method is performed in two main phases. First, the segmentation units of the input multi-spectral RS image data are subjected to objectbased interval-valued symbolic modeling. As a result, features for each land cover type are represented in the form of an intervalvalued symbolic vector, which describes the intra-class uncertainty better than the source data and improves the separability between different classes. Second, interval type-2 fuzzy sets are generated for each cluster based on the distance metric of the interval-valued vectors. This step characterizes the inter-class high-order fuzzy uncertainty and improves the classification accuracy. To demonstrate the advantages and effectiveness of the proposed approach, extensive experiments are conducted on two multispectral RS image datasets from regions with complex land cover characteristics, and the results are compared with those given by well-known fuzzy and conventional clustering algorithms.     

High-resolution remote sensing mapping of global land water

Land water, one of the important components of land cover, is the indispensable and important basic information for climate change studies, ecological environment assessment, macro-control analysis, etc. This article describes the overall study on land water in the program of global land cover remote sensing mapping. Through collection and processing of Landsat TM/ETM+, China’s HJ-1 satellite image, etc., the program achieves an effective overlay of global multi-spectral image of 30 m resolution for two base years, namely, 2000 and 2010, with the image rectification accuracy meeting the requirements of 1:200000 mapping and the error in registration of images for the two periods being controlled within 1 pixel. The indexes were designed and selected reasonably based on spectral features and geometric shapes of water on the scale of 30 m resolution, the water information was extracted in an elaborate way by combining a simple and easy operation through pixel-based classification method with a comprehensive utilization of various rules and knowledge through the object-oriented classification method, and finally the classification results were further optimized and improved by the human-computer interaction, thus realizing high-resolution remote sensing mapping of global water. The completed global land water data results, including Global Land 30-water 2000 and Global Land 30-water 2010, are the classification results featuring the highest resolution on a global scale, and the overall accuracy of self-assessment is 96%. These data are the important basic data for developing relevant studies, such as analyzing spatial distribution pattern of global land water, revealing regional difference, studying space-time fluctuation law, and diagnosing health of ecological environment.     

Land use and land cover classification using phenological variability from MODIS vegetation in the Upper Pangani River Basin,Eastern Africa

In arid and semi-arid areas, evaporation fluxes are the largest component of the hydrological cycle, with runoff coefficient rarely exceeding 10%. These fluxes are a function of land use and land management and as such an essential component for integrated water resources management. Spatially distributed land use and land cover (LULC) maps distinguishing not only natural land cover but also management practices such as irrigation are therefore essential for comprehensive water management analysis in a river basin. Through remote sensing, LULC can be classified using its unique phenological variability observed over time. For this purpose, sixteen LULC types have been classified in the Upper Pangani River Basin (the headwaters of the Pangani River Basin in Tanzania) using MODIS vegetation satellite data. Ninety-four images based on 8 day temporal and 250 m spatial resolutions were analyzed for the hydrological years 2009 and 2010. Unsupervised and supervised clustering techniques were utilized to identify various LULC types with aid of ground information on crop calendar and the land features of the river basin. Ground truthing data were obtained during two rainfall seasons to assess the classification accuracy. The results showed an overall classification accuracy of 85%, with the producer’s accuracy of 83% and user’s accuracy of 86% for confidence level of 98% in the analysis. The overall Kappa coefficient of 0.85 also showed good agreement between the LULC and the ground data. The land suitability classification based on FAO-SYS framework for the various LULC types were also consistent with the derived classification results. The existing local database on total smallholder irrigation development and sugarcane cultivation (large scale irrigation) showed a 74% and 95% variation respectively to the LULC classification and showed fairly good geographical distribution. The LULC information provides an essential boundary condition for establishing the water use and management of green and blue water resources in the water stress Pangani River Basin.     

Estimating leaf area index by inversion of reflectance model for semiarid natural grasslands

The study developed an integrated reflectance model combining radiative transfer and geometric optical properties in order to inverse leaf area index (LAI) of semiarid natural grasslands. In order to better link remote sensing information with land plants, and facilitate regional and global climate change studies, the model introduced a simple but important geometrical similarity parameter related to plant crown shapes. The model revealed the influences of different plant crown shapes (such as spherical, cylindrical/cuboidal and conic crowns) on leaf/branch angle distribution frequencies, shadow ground coverage, shadowed or sunlit background fractions, canopy reflectance, and scene reflectance. The modeled reflectance data agreed with the measured ones in the three Leymus chinensis steppes with different degradation degrees, which validated the reflectance model. The lower the degradation degree was, the better the modeled data agreed with the measured data. After this reflectance model was coupled with the optimization inversion method, LAI over the entire study region was estimated once every eight days using the eight-day products of surface reflectance obtained by multi-spectral Moderate-Resolution Imaging Spectroradiometer (MODIS) during the growing seasons in 2002. The temporal and spatial patterns of inversed LAI for the steppes with different cover degrees, swamps, flood plains, and croplands agreed with the general laws and measurements very well. But for unused land cover types (sands, saline, and barren lands) and forestlands, totally accounting for about 10% of the study region, the reasonable LAI values were not derived by inversing, requiring further revising of the model or the development of a new model for them. Supported by the National Natural Science Foundation of China (Grant No. 30500076), the National Basic Research Program of China (Grant No. 2007CB714407), and the President Foundation of Graduate University of Chinese Academy of Sciences (Grant No. YZJJ200205)     

Improvement of ecological geographic regionalization based on remote sensing and canonical correspondence analysis: A case study in China

Ecological geographic regions, also called eco-regions, can be used to divide a remotely sensed image, which is helpful for reducing the complexity of land cover types within eco-regions and for improving the classification accuracy of land cover. In this case study in China, we improved a method of ecological geographic regionalization that is more suitable for remote sensing mapping of regional land cover, and we obtained new eco-regions. The canonical correspondence analysis (CCA) and k-means clustering were adopted in the ecological geographic regionalization using both seasonal remotely-sensed vegetation information and environmental data including climate, elevation and soil features. Our results show that the combination of seasonal vegetation information and the CCA performed well in the selection of the dominant environmental factor of the biogeographic pattern, and it can be used as regionalization indicators of eco-regions. We found that thermal factors are the most important driving forces of the biogeographic pattern in China, which followed by moisture factors. Two global land cover products (MODIS MCD12C1 and GlobCover) were used to assess our eco-regions. The results show that our eco-regions performed better than that of a previous study regarding the complexity of land cover types, such as in the number of types and the proportional area of the major/secondary type. These results indicate that the method of ecological geographic regionalization, which is based on environmental factors associated with seasonal vegetation features, is effective for reducing the regional complexity of land cover.     

近岸Ⅱ类水体表层悬浮泥沙浓度遥感模式研究进展

因为具有明显的时间与空间分辨率优势,遥感数据成为近岸Ⅱ类水体悬浮泥沙浓度(SSC)信息提取研究的重要数据源之一.悬浮泥沙遥感信息提取的现状可归纳为:(1)建立近岸Ⅱ类水体SSC遥感模式的方法有三种类型,分别是基于地面光谱与SSC测量的反射率反演方法、基于图像信息法和基于大气辐射传输理论模型法;(2)基于地面测量的反射率反演方法属于理论与经验相结合的方法,也是目前用于SSC定量化遥感模式研究的常用方法.其数学表达形式包括线性关系式、对数关系式、负指数关系式、Gordon模式和综合模式等;(3)到目前为止已有的Ⅱ类水体SSC遥感模式适用性方面还不理想,远未达到与试验室分析相匹配的精度.文章认为:加强地面水文光谱实验研究,建立多光谱SSC定量模式,以高分辨率和高光谱遥感融合数据为基础的SSC定量遥感是今后该方向发展趋势.     

Modelling the water balance of a mesoscale catchment basin using remotely sensed land cover data

Hydrological modelling of mesoscale catchments is often adversely affected by a lack of adequate information about specific site conditions. In particular, digital land cover data are available from data sets which were acquired on a European or a national scale. These data sets do not only exhibit a restricted spatial resolution but also a differentiation of crops and impervious areas which is not appropriate to the needs of mesoscale hydrological models. In this paper, the impact of remote sensing data on the reliability of a water balance model is investigated and compared to model results determined on the basis of CORINE (Coordination of Information on the Environment) Land Cover as a reference. The aim is to quantify the improved model performance achieved by an enhanced land cover representation and corresponding model modifications. Making use of medium resolution satellite imagery from SPOT, LANDSAT ETM+ and ASTER, detailed information on land cover, especially agricultural crops and impervious surfaces, was extracted over a 5-year period (2000–2004). Crop-specific evapotranspiration coefficients were derived by using remote sensing data to replace grass reference evapotranspiration necessitated by the use of CORINE land cover for rural areas. For regions classified as settlement or industrial areas, degrees of imperviousness were derived. The data were incorporated into the hydrological model GROWA (large-scale water balance model), which uses an empirical approach combining distributed meteorological data with distributed site parameters to calculate the annual runoff components. Using satellite imagery in combination with runoff data from gauging stations for the years 2000–2004, the actual evapotranspiration calculation in GROWA was methodologically extended by including empirical crop coefficients for actual evapotranspiration calculations. While GROWA originally treated agricultural areas as homogeneous, now a consideration and differentiation of the main crops is possible. The accuracy was determined by runoff measurements from gauging stations. Differences in water balances resulting from the use of remote sensing data as opposed to CORINE were analysed in this study using a representative subcatchment. Resulting Nash–Sutcliff model efficiencies improved from 0.372 to 0.775 and indicate that the enhanced model can produce thematically more accurate and spatially more detailed local water balances. However, the proposed model enhancements by satellite imagery have not exhausted the full potential of water balance modelling, for which a higher temporal resolution is required.     

Modeling the land surface reflectance for optical remote sensing data in rugged terrain

A model for topographic correction and land surface reflectance estimation for optical remote sensing data in rugged terrian is presented.Considering a directional-directional reflectance that is used for direct solar irradiance correction and a hemispheric-directional reflectance that is used for atmospheric diffuse irradiance and terrain background reflected irradiance correction respectively,the directional reflectance-based model for topographic effects removing and land surface reflectance calculation is developed by deducing the directional reflectance with topographic effects and using a radiative transfer model.A canopy reflectance simulated by GOMS model and Landsat/TM raw data covering Jiangxi rugged area were taken to validate the performance of the model presented in the paper.The validation results show that the model presented here has a remarkable ability to correct topography and estimate land surface reflectance and also provides a technique method for sequently quantitative remote sensing application in terrain area.     

Object-oriented land cover classification using HJ-1 remote sensing imagery

The object-oriented information extraction technique was used to improve classification accuracy, and addressed the problem that HJ-1 CCD remote sensing images have only four spectral bands with moderate spatial resolution. We used two key techniques: the selection of optimum image segmentation scale and the development of an appropriate object-oriented information extraction strategy. With the principle of minimizing merge cost of merging neighboring pixels/objects, we used spatial autocorrelation index Moran’s I and the variance index to select the optimum segmentation scale. The Nearest Neighborhood (NN) classifier based on sampling and a knowledge-based fuzzy classifier were used in the object-oriented information extraction strategy. In this classification step, feature optimization was used to improve information extraction accuracy using reduced data dimension. These two techniques were applied to land cover information extraction for Shanghai city using a HJ-1 CCD image. Results indicate that the information extraction accuracy of the object-oriented method was much higher than that of the pixel-based method.     

基于实测光谱数据的太湖水华和水生高等植物识别

水华和水生高等植物的识别对于内陆水质遥感监测至关重要,其分布状况既可以用于表征水牛植物的分布状况,从而间接反映水质分布状况,又可以进一步利用非水华和水草的水体遥感数据进行水质参数反演.然而,常用的多光谱遥感数据很难精确识别水华和水草,只有高光谱遥感数据才能够捕捉水华、水草和水体细致的光谱差异,从而对水华和水草进行精确识别.但是目前还缺乏基于高光谱遥感数据的水华和水草识别的系统性研究.以太湖为研究区,于2006年7月和10月开展了2次水面光谱测量实验,获取了水华、浮叶植物、沉水植物和水体的反射率光谱.在光谱分析的基础上,首先建它了4种光谱指数,进而利用这4种光谱指数建立了水华、浮叶植物、沉水植物和水体的判别公式,并利用2006年10月水面实验测量的光谱数据训练得到了判别公式中的阈值.通过2006年7月水面实验测量的光谱数据的检验,证明提出的判别公式能够很好的识别水华和水草,获得了较高的识别精度.     

Remote Sensing Data Binary Classification Using Boosting with Simple Classifiers

Boosting is a classification method which has been proven useful in non-satellite image processing while it is still new to satellite remote sensing. It is a meta-algorithm, which builds a strong classifier from many weak ones in iterative way. We adapt the AdaBoost.M1 boosting algorithm in a new land cover classification scenario based on utilization of very simple threshold classifiers employing spectral and contextual information. Thresholds for the classifiers are automatically calculated adaptively to data statistics.The proposed method is employed for the exemplary problem of artificial area identification. Classification of IKONOS multispectral data results in short computational time and overall accuracy of 94.4% comparing to 94.0% obtained by using AdaBoost.M1 with trees and 93.8% achieved using Random Forest. The influence of a manipulation of the final threshold of the strong classifier on classification results is reported.     

An assessment of water reserve changes in Salt Lake,Turkey, through multi‐temporal Landsat imagery and real‐time ground surveys

This paper focuses mainly on the investigation of water reserve changes in Salt Lake, Turkey, using remote‐sensing data. The study is performed in two stages: (1) correlation analysis for real‐time ground and satellite data and (2) assessment of water reserve changes using multi‐temporal Landsat imagery. First, correlation analysis is conducted to investigate the relationship between digital data from Landsat‐5 TM and spectral (in situ) measurements collected using a field spectroradiometer on the same day and time. A radiometric correction procedure, including conversions from digital numbers to radiance and from radiance to at‐satellite reflectance, is executed to make satellite data comparable to in situ measurements. This procedure show that simultaneous ground and satellite remote‐sensing data are highly correlated (0·84 > R² > 97) and the near‐infrared region (for this study TM4‐Landsat‐5 TM, band 4) is the best spectral range to distinguish salt and water on the satellite data for the multi‐temporal analysis of the water reserve in Salt Lake. It also shows that the use of shortwave infrared band(s) will result in confusion for the determination of the water reserve in this water‐covered study area. In a second and last phase, the water reserve change in the lake is examined using multi‐temporal Landsat imagery collected in 1990, 2001 and 2005. The remotely sensed, sampled and treated data show that the water reserve in the lake has decreased markedly between 1990 and 2005 due to drought and uncontrolled water usage. It is suggested that the use of water supplies around Salt Lake should be controlled and that the lake should regularly be monitored by up‐to‐date remote‐sensing data (at least annually) for better management of water resources. Copyright © 2007 John Wiley & Sons, Ltd.     

查干湖冰封期光谱特征及影响因素

湖冰光谱特征是湖冰遥感反演的物理基础,是研究湖冰光学特性和空间分布的理论依据。本文以查干湖为例,使用ASD Field Spec 4便携式地物光谱仪采集冰封期不同类型湖冰、积雪和水体光谱,利用Savitzky-Golay滤波法和包络线去除法分析白冰、灰冰、黑冰、雪冰、积雪和水体的反射光谱特征,探索气泡对湖冰反射光谱特征的影响。积雪和雪冰、白冰和灰冰、黑冰和水体的反射特征随着波长的变化特征基本一致,冰的反射率介于积雪和水体之间,其中白冰的反射率高于灰冰和黑冰,在包络线去除结果中,黑冰和水体在440 nm吸收谷处的吸收面积为5.184和10.878、吸收深度为0.052和0.106,雪、雪冰、白冰、灰冰在800和1030 nm吸收谷处的吸收面积和吸收深度的变化表现为雪<雪冰<灰冰<白冰。气泡是影响湖冰光谱特征的重要因素,气泡使白冰反射率减小和黑冰反射率增大,并且气泡使得白冰在800/1030nm和黑冰在440 nm处的吸收面积和吸收深度减小,其中气泡大小和疏密程度的不同会导致湖冰反射率的影响程度存在差异。同时,本文选取时间同步的Landsat 8 OLI遥感影像,在完成辐...     

Land cover mapping using time series HJ-1/CCD data

It is very difficult to have remote sensing data with both high spatial resolution and high temporal frequency; thus, two categories of land-use mapping methodology have been developed separately for coarser resolution and finer resolution data. The first category uses time series of data to retrieve the variation of land surface for classification, which are usually used for coarser resolution data with high temporal frequency. The second category uses fine spatial resolution data to classify different land surface. With the launch of Chinese satellite constellation HJ-1in 2008, four 30 m spatial resolution CCDs with about 360 km coverage for each one onboard two satellites made a revisit period of two days, which brought a new type of data with both high spatial resolution and high temporal frequency. Therefore, by taking the spatiotemporal advantage of HJ-1/CCD data we propose a new method for finer resolution land cover mapping using the time series HJ-1/CCD data, which can greatly improve the land cover mapping accuracy. In our two study areas, the very high resolution remote sensing data within Google Earth are used to validate the land cover mapping results, which shows a very high mapping accuracy of 95.76% and 83.78% and a high Kappa coefficient of 0.9423 and 0.8165 in the Dahuofang area of Liaoning Province and the Heiquan area of Gansu Province respectively.     

Optimum segmentation of simple objects in high-resolution remote sensing imagery in coastal areas

The optimum segmentation of ground objects in a landscape is essential for interpretation of high-resolution remotely sensed imagery and detection of objects; and it is also a technical foundation to efficiently use spatial information in remote sensing imagery. Landscapes are complex system composed of a large number of heterogeneous components. There are many explicit homogeneous image objects that have similar spectral character and yet differ from surrounding objects in high-resolution remote sensing imagery. Thereby, a new concept of Distinctive Feature of fractal is put forward and used in deriving Distinctive Feature curve of fractal evolution in multiscale segmentation. Through distinguishing the extremum condition of Distinctive Feature curve and the inclusion relationship of fractals in multiscale representation the Scalar Order is built. This can help to determinate the optimum scale in image segmentation for simple-objects, and the potential meaningful image-object fitting the intrinsic scale of the dominant landscape object can be obtained. Based on the application in high-resolution remote sensing imagery in coastal areas, a satisfactory result was acquired.     

Surveying earth resources by remote sensing from satellites

This paper reviews the techniques and recent results of orbital remote sensing, with emphasis on Landsat and Skylab imagery. Landsat (formerly ERTS) uses electronic sensors (scanners and television) for repetitive observations with moderate ground resolution. The Skylab flights used a wider range of electro-optical sensors and returned film cameras with moderate and high ground resolution. Data from these programs have been used successfully in many fields. For mineral resources, satellite observations have proven valuable in geologic mapping and in exploration for metal, oil, and gas deposits, generally as a guide for other (conventional) techniques. Water resource monitoring with satellite data has included hydrologic mapping, soil moisture studies, and snow surveys. Marine resources have been studied, with applications in the fishing industry and in ocean transportation. Agricultural applications, benefiting from the repetitive coverage possible with satellites, have been especially promising. Crop inventories are being conducted, as well as inventories of timber and rangeland. Overgrazing has been monitored in several areas. Finally, environmental quality has also proven susceptible to orbital remote sensing; several types of water pollution have been successfully monitored. The effects of mining and other activities on the land can also be studied. The future of orbital remote sensing in global monitoring of the Earth's resources seems assured. However, efforts to extend spectral range, increase resolution, and solve cloud-cover problems must be continued. Broad applications of computer analysis techniques are vital to handle the immense amount of information produced by satellite sensors.Abbreviations SSRERTS Symposium on Significant Results Obtained from the Earth Resources Technology Satellite - TERTSS Third Earth Resources Technology Satellite-1 Symposium National Academy of Sciences-National Research Council Research Associate at NASA Goddard Space Flight Center, 1974/75.Goddard Space Flight Center     

Assessing Ground Water Development Potential Using Landsat Imagery

Seven villages in southeastern Kenya surround Mt. Kasigau and depend on the mountain's cloud forest for their water supply. Five of these villages have regularly experienced water shortages, and all village water supplies were contaminated with Escherichia coli bacteria. There is a need to economically find new sources of fresh ground water. Remote sensing offers a relatively quick and cost-effective way of identifying areas with high potential for ground water development. This study used spectral properties of features on Landsat remote sensing imagery to map linear features, soil types, surface moisture, and vegetation. Linear features represented geologic or geomorphologic features indicating either shallow ground water or areas of increased subsurface hydraulic conductivity. Regarding soil type, black soils were identified as potential indicators of shallow aquifers based on their relatively lower elevation and association with river valleys. A vegetation map was created using unsupervised classification, and three of the resulting vegetation classes were observed to be commonly associated with wet areas and/or ground water discharge. A wetness map, created using tasseled cap analysis, was used to identify all areas of high ground moisture, including those that corresponded to vegetated areas. The linear features, soil type, vegetation, and wetness maps were overlaid to produce a composite that highlighted areas with the highest potential for ground water development. Electrical resistivity surveys confirmed that areas highlighted by the composite image had relatively shallow depths to the water table. Some figures in this paper are available in color in the online version of the paper.     

基于Lasso函数的分层混合模糊-神经网络及其在遥感影像分类中的应用

本文提出一种新的分层混合模糊-神经网络(HHFNN)算法.在模糊系统中使用Takagi-Sugeno模型和三角波隶属函数.同时,为降低离散输入变量中可能存在的强交瓦作用,采用了系数收缩机制中的Lasso函数.最后,以福建的漳平洛阳—安溪潘田地区LANDSAT ETM+遥感影像数据地物分类为例,应用本文的改进算法与其他神...     

A CASE STUDY ON RULE-BASED MULTIPLE SOURCE DATA CLASSIFICATION WITH GIS

IINTRODUCTIONLanduse/coverisoneofthemostimportalfactorseffectingsoilandwaterloss.Researchonlanduse/coverandwaterandsoillosswithremotesensinghasbeendonemuchinthepast.Butremotelysensedimageryonlycarriestheinstantaneousandtwo-dimensionalinformationofitsprototypegeographicobjects(ChenandZhao,1990).Thereforemathematicalandphysicalprocessingonremotelysensedimageoftenproduceindefiniteandunreliableresult.Inordertoimprovetheprecisionofclassification,otherdatasetssuchastopographicmaps,thematicmaps…