增强型DeepLab算法和自适应损失函数的高分辨率遥感影像分类

高分辨率遥感影像地物复杂,分类难度大,而深度学习方法可以提取地物更多更深层次的特征信息,适用于高分辨率遥感影像的地物分类。本文研究对高分辨率影像中不透水地面、建筑、低矮植被、树、车辆等地物的高精度分类。结合遥感多地物分类的特点,以DeepLab v3+网络模型为基础,提出E-DeepLab网络模型。主要改进为：（1）改进编码器和解码器的结合方式,使用简洁有效的加成连接方式。（2）缩小单次上采样倍数,增加上采样层,提高编码器与解码器连接的紧密性。（3）使用改进的自适应权重损失函数,自动调节地物损失权重。同时根据数据特点,提出结合DSM、NDVI数据等多通道训练方式。使用两个地区数据进行实验,结果表明,两地区精度均明显优于原始DeepLab v3+模型和其他相关模型,Potsdam地区总体提取精度达到93.2%,建筑物提取精度达到97.8%,Vaihingen地区总体提取精度达到90.7%,建筑物提取精度达到96.3%。目视对比分类图和标准标记图,两者具有高度的一致性。本文所提出的E-DeepLab网络在高分辨率遥感影像地物高精度提取和分类中有较好的应用价值。     

遥感图像分类精度的点、群样本检验与评估

遥感专题分类结果在使用前,必须进行客观可靠的精度验证和分析,以保持遥感分类结果的可靠性.本文利用不同分辨率遥感数据获取的同一地区土地利用/覆盖信息,进行了简单随机抽样、系统抽样和分层抽样三种不同抽样组织方式下的点样本和群样本检验分析,评估了不同抽样方式下的点样本和群样本检验效果.研究结果表明：（1）抽样方式对遥感分类精度评价结果的影响是客观存在的,不同抽样方式下的点样本和群样本检验结果都存在一定的随机性,但同一种抽样方式下,点样本检验精度评估结果的波动幅度小于群样本检验,稳定性比群样本检验要好;（2）不同抽样方式下的多次点样本和群样本检验的平均精度检验结果基本上都能够反映分类图像的精度特征,其中,点样本检验中,分层随机抽样点样本检验效果较好;群样本检验中,系统抽样群样本检验和分层随机抽样群样本检验的效果优于简单随机抽样群样本检验.     

基于证据理论的水稻遥感信息提取方法研究

水稻播种面积是农业管理部门关心的重要问题之一。本文介绍了证据理论在遥感图像分类上的应用,并以汉川市为示范区,采用2002年TM遥感影像,在GIS技术支持下,通过建模运算,剔除不可能是水稻的像元。然后将证据理论用于影像分类,将影像初次分类结果,与参考图对照,将不满足要求的区域提取出来,再次进行第二次分类。将第二次分类结果与参考图对照,显示分类效果满足要求;若不满足要求,可继续进行再次分类,直到分类效果满意为止。将两次分类结果中的水稻信息进行叠加,提取出水稻遥感信息,经检验精度达到91.39%。     

What is the difference between two maps? A remote senser’s view

In remote sensing, thematic map comparison is often undertaken on a per-pixel basis and based upon measures of classification agreement. Here, the degree of agreement between two thematic maps, and so the difference between the pair, was evaluated through visual and quantitative analyses for two scenarios. Quantitative assessments were based on basic site-specific measures of agreement that are used widely in accuracy assessment (e.g. the overall percentage of pixels with the same class label in each of the two maps and the kappa coefficient of agreement) as well as an information theory based approach that allows the degree of mutual or shared information to be assessed even if different classification schemes have been used to produce the maps. The results indicated that in the first map comparison scenario, focused on labelling, there was a fair degree of correspondence between the maps but with an overall difference in information content of ∼42%. In the second comparison scenario, focused on change in time, considerable change had occurred with a change in class label for ∼42% of the pixels. It was also apparent that global assessments masked local scale changes.     

Toward accountable land use mapping: Using geocomputation to improve classification accuracy and reveal uncertainty

The classification of satellite imagery into land use/cover maps is a major challenge in the field of remote sensing. This research aimed at improving the classification accuracy while also revealing uncertain areas by employing a geocomputational approach. We computed numerous land use maps by considering both image texture and band ratio information in the classification procedure. For each land use class, those classifications with the highest class-accuracy were selected and combined into class-probability maps. By selecting the land use class with highest probability for each pixel, we created a hard classification. We stored the corresponding class probabilities in a separate map, indicating the spatial uncertainty in the hard classification. By combining the uncertainty map and the hard classification we created a probability-based land use map, containing spatial estimates of the uncertainty. The technique was tested for both ASTER and Landsat 5 satellite imagery of Gorizia, Italy, and resulted in a 34% and 31% increase, respectively, in the kappa coefficient of classification accuracy. We believe that geocomputational classification methods can be used generally to improve land use and land cover classification from imagery, and to help incorporate classification uncertainty into the resultant map themes.     

高分辨率遥感影像的压缩纹理元分类

针对传统的高分辨率遥感影像分类中特征提取复杂,特征维数大等问题,提出一种新颖,简单,高效的纹理特征提取方法。首先,利用随机投影对基于原始像素灰度值的纹理元矢量进行降维,将其投影到压缩的纹理特征子空间。然后,在压缩子空间中对各类纹理元进行聚类,将聚类中心作为纹理字典,得到局部纹理特征集。最后,将样本中包含的纹理元编码到纹理字典中对应距离最近的词汇,得到样本的视觉词汇图,并融合词汇统计直方图与词汇二阶矩信息作为最终的纹理表达。通过两组实验,验证了本文方法能够有效的表达纹理,提高分类精度。     

典型遥感数据分类方法的适用性分析——以遥感图像场景分类为例

目前普遍采用的分类器通常都是针对单一或小量任务而设计的,在小数据量的处理中能取得比较满意的结果。但对于海量遥感数据的处理,其在处理时效和分类精度方面还有待研究。本文以遥感图像场景分类任务为例,着重对遥感数据分类问题中几种典型分类方法的适用性进行比较研究,包括K近邻(KNN)、随机森林(RF),支持向量机(SVM)和稀疏表达分类器(SRC)等。分别从参数敏感性,训练样本数据量,待分类样本数据量和样本特征维数对分类器性能的影响等几个方面进行比较分析。实验结果表明:(1)KNN,RF和L0-SRC方法相比RBF-SVM,Linear-SVM和L1-SRC,受参数影响的程度更弱;(2)待分类样本固定的情况下,随着训练样本数目的增加,SRC类型分类方法的分类性能最佳,SVM类型方法次之,然后是RF和KNN,在总体分类时间上呈现出L0-SRCL1-SRCRFRBF-SVM/Linear-SVMKNN/L0-SRC-Batch的趋势;(3)训练样本固定的情况下,所有分类方法的分类精度几乎都不受待分类样本数目变化的影响,RBF-SVM方法性能最佳,其次是L1-SRC,然后是Linear-SVM,最后是RF和L0-SRC/L0-SRC-Batch,在总体分类时间上,L1-SRC和L0-SRC相比其他分类方法最为耗时;(4)样本特征维数的变化不仅影响分类器的运行效率,同时也影响其分类精度,其中SRC和KNN分类器器无需较高的特征维数即可获得较好的分类结果,SVM对高维特征具有较强的包容性和学习能力,RF分类器对特征维数增加则表现得并不敏感,特征维数的增加并不能对其分类精度的提升带来更多的贡献。总的来说,在大数据量的遥感数据分类任务中,现有分类方法具有良好的适用性,但是对于分类器的选择应当基于各自的特点和优势,结合实际应用的特点进行权衡和选择,选择参数敏感性较小,分类总体时间消耗低但分类精度相对较高的分类方法。     

分层神经网络分类算法

提高遥感图像分类精度一直是爱到普遍关注的焦点问题。近年来,人工神经网络技术和分 处理技术由于它们的许多优点受到广泛欢迎。本文把两种技术结合起来,提出了分层神经网络的概念,并基于此设计了一种分层神经网络分类算法。通过与最大似然法的对比实验表明,这种分层神经网络分类算法可以明显地提高分类精度,并对不规则分布的复杂数据具有很强的处理能力。     

The UK land cover map 2000: Planning,construction and calibration of a remotely sensed,user-oriented map of broad habitats

Land cover map 2000 (LCM2000) is a comprehensive survey of UK broad habitats giving vector digital maps from segment-based classification of remotely sensed satellite data. This paper examines the influence of users in designing LCM2000 and the difficulties in applying a user-defined classification. It assesses problems and successes through comparisons with a sample-based field survey. These suggest that LCM2000 accuracy at broad habitat level may be around 80–85%; however, it was not possible fully to discriminate errors in LCM2000 from those of the field survey or from mismatches in scales, resolutions and survey dates. Calibration generated broad habitat cover statistics from LCM2000 data to field survey equivalence. These take full account of the heterogeneity of a study area, helping to generate accurate statistics, including those at local level where the field survey cannot operate effectively. The paper concludes that the comprehensive and extensive coverage from remote sensing comes closer than alternative methods to meeting users needs. However, it recognises that producers of remotely sensed information need to understand better the needs of users, and users need to appreciate what the technology can and cannot deliver. This paper adds some benefits of hindsight to the process of communication.     

Development of S-NPP VIIRS global surface type classification map using support vector machines

With the launch of the Joint Polar Satellite System (JPSS)/Suomi National Polar-orbiting Partnership (S-NPP) satellite in October 2011, many of the terrestrial remote sensing products generated from Moderate Resolution Imaging Spectroradiometer (MODIS), such as the global land cover map, have been inherited and expanded into the JPSS/S-NPP mission using the new Visible Infrared Imaging Radiometer Suite (VIIRS) data. In this study, an improved algorithm including the use of a new classifier support vector machines (SVM) classifier was proposed to produce VIIRS surface type maps. In addition to the new classification algorithm, a new post-processing strategy involving the use of new ancillary data to refine the classification output is implemented. As a result, the new global International Geosphere-Biosphere Programme (IGBP) map based on the 2014 VIIRS surface reflectance data was generated with a 78.5 ± 0.6% overall classification accuracy. The new map was compared to a previously delivered VIIRS surface type map, and to the MODIS land cover product. Validation results including the error matrix, overall accuracy, and the user’s and producer’s accuracy suggest the new global surface type map provides similar classification accuracy compared to the old VIIRS surface type map, with higher accuracy achieved in agricultural types.     

西藏矿产资源潜力评价遥感专题中典型问题初探

西藏矿产资源潜力评价项目遥感专题通过提取各1:25万标准图幅及全区遥感异常信息和编制遥感推断地质构造图,为项目地质背景编图、成矿规律研究和矿产资源预测提供了基础性图件.因西藏全区面积广阔、涉及的遥感图像时相各异、图像处理软件及方法流程复杂,加之技术人员专业背景和经验等差异,在遥感专题实施中出现了4个典型问题:①对基准面...     

基于贝叶斯网络分类的遥感影像变化检测

遥感成像过程中,地面、大气等诸多要素的不确定性和波段之间的相关性等原因影响了分类精度,导致变化检测的不准确性。为了提高分类精度往往需要引入先验知识。贝叶斯网络是一种新的数据表达和推理模型,对数据没有严格的正态分布前提要求,通过动态地调整先验概率密度,能有效提高分类精度。以北京通州地区1996-05-29和2001-05-19两个时相的陆地卫星Landsat TM遥感影像为例,介绍了基于贝叶斯网络的分类算法,并在此基础上实现了两个时相遥感影像的变化检测。实验结果表明:基于贝叶斯网络分类算法的后分类比较变化检测方法是遥感影像变化检测的一种新的有效方法。     

Categorical maps, comparisons, and confidence

The comparison of categorical maps is a common problem in several different contexts. Differences between categorical maps can be characterized and measured in a variety of ways. In 2004 we invited individuals from remote sensing, geographical information analysis, spatial modelling, and landscape ecology to participate in a virtual workshop in order to compare strategies for comparison. This revealed that the key dimensions of comparison relate to (1) the map characteristics considered, (2) the nature of the comparison, (3) the handling of geographical space, (4) the units of computed measures, and (5) the tests of significance.Ferko Csillag deceased     

Mapping dominant vegetation communities at Meili Snow Mountain,Yunnan Province,China using satellite imagery and plant community data

Mapping dominant vegetation communities is important work for vegetation scientists. It is very difficult to map dominant vegetation communities using multispectral remote sensing data only, especially in mountain areas. However plant community data contain useful information about the relationships between plant communities and their environment. In this paper, plant community data are linked with remote sensing to map vegetation communities. The Bayesian soft classifier was used to produce posterior probability images for each class. These images were used to calculate the prior probabilities. One hundred and eighty plant plots at Meili Snow Mountain, Yunnan Province, China were used to characterize the vegetation distribution for each class along altitude gradients. Then, the frequencies were used to modify the prior probabilities of each class. After stratification in a vegetation part and a non-vegetation part, a maximum-likelihood classification with equal prior probabilities was conducted, yielding an overall accuracy of 82.1% and a kappa accuracy of 0.797. Maximum-likelihood classification with modified prior probabilities in the vegetation part, conducted with a conventional maximum-likelihood classification for the non-vegetation part, yielded an overall accuracy of 87.7%, and a kappa accuracy of 0.861.     

土壤调查地区级汇总的遥感制图试验研究

本研究的目的是探讨利用遥感影像数据,对县级土壤调查资料进行地区级汇总的制图工作方法。我们依据的制图综合原则是:(1)以现有的土壤图为基础,充分尊重原图上的界线,主要以土层界线重新综合;(2)研究并搞清各种土壤分布规律和它们之间的组合规律,然后进行归并和区分;(3)参照卫星影像特征和判读标志,对某些明显与实地不符之处,以影像和实地调查资料进行修正,有些土壤类型以组合形式表达。本文对利用遥感数据进行制图综合的工作方法进行了详细描述,其中包括:制图单元的综合和比例尺的改变,底图的制作和利用卫星影像对综合后的土壤图进行编制等。研究结果表明,此种方法与常规方法比较,特别是对有一定物质和技术条件的单位(如省级),不仅在土壤调查中,而且在资源调查中皆可采用此种编图技术,收到节省人力、物力和时间的效果。     

结合L0平滑和超像素的高空间遥感影像非监督分类

针对基于像元的非监督分类方法对高空间遥感影像分类时易形成“椒盐”噪声和产生大量错分、漏分的问题,提出了一种结合L0平滑和超像素的非监督分类方法．首先采用L0算法对高空间遥感影像进行平滑操作,减少大量图像噪声及冗余信息;然后采用简单的线性迭代聚类（SLIC）超像素方法处理平滑后图像,进一步抑制椒盐现象的同时降低处理复杂度,得到初始聚类图;最后采用K-means非监督分类方法得到最终分类结果图．为验证本文提出的方法,选取3景高空间遥感影像作为实验数据．试验结果表明,采用提出的方法能准确对地物分类,且总体精度分别达到了72.46％、77.55％和78.44％,Kappa系数分别达到0.788、0.779和0.779．提出方法能有效解决分类中存在的“椒盐”现象,可提高分类精度,对高空间遥感影像分类具有一定的参考价值．      

SVM多窗口纹理土地利用信息提取技术

针对单一窗口纹理分类时地物破碎,分类精度不高等问题,提出了一种基于支持向量机多窗口纹理的遥感图像分类方法。该方法在对SPOT5遥感影像进行纹理特征提取的基础上,构建了结合多窗口纹理的SVM模型。以陕西省佛坪县长角坝乡为试验区,利用此模型对该区域的土地利用类型进行分类研究,并将分类结果与单一窗口纹理SVM分类和单元数据(光谱)SVM分类结果进行了比较分析。结果表明:多窗口纹理参与的土地利用分类总精度达到85.33%,比单一窗口纹理分类提高了13.11%,而与单元数据SVM分类相比提高了近24.10%,取得了较好的分类效果,有效地解决了单一窗口纹理分类时地物破碎、分类精度不高等问题。     

A History and Analysis of Mapping Urban Expansion in the Kathmandu Valley,Nepal

Abstract

There have been numerous efforts over many years to map or delineate urban locations and features in the Kathmandu Valley of Nepal. This study acquired 27 land useland cover maps for the Valley or the urban portions of the Valley. Those maps vary greatly in their mapping parameters. The objectives of this study were to first conduct a cartographic comparison of the differences in the creation and content of the maps and then do an analysis of the urban changes in the Valley based upon the maps. The maps for the Valley have differed in their source materials, the amount of field work involved, scale and minimum mapping units, classifications used, definitions of classes and coordinate systems. Source materials have included various scales and formats of aerial photography and different satellite systems. The most difficult issue in comparing the maps is the varied classification systems and definitions. The same feature will be classified differently from map to map. This is particularly an issue for institutional features such as temples, palaces, educational facilities, open public space and governmental sites. Definitions of residential areas are also not consistent. Even with the differences in mapping parameters, considerable useful information can be obtained by comparing these maps. These include a simply documentation of the urban extent and the generally resulting loss in agricultural lands. There was an increase in urban extent from 22 to 83km² between 1955 and 2000. Urban expansion has also changed from occurring on the upland river terraces or tars to the floodplains. Finally, while not directly documented in these maps, the tremendous pace of urban growth has resulted in multiple infrastructure and environmental challenges.     

基于分类规则挖掘的遥感影像分类研究

分析了目前遥感影像的统计分类、神经网络分类及基于符号知识的逻辑推理分类方法的优缺点.以GIS为平台,构建了多源空间数据库,将数据挖掘的思想和方法引入遥感影像分类中,提出了面向分类规则挖掘的遥感影像分类框架.针对遥感光谱数据及其他空间数据的特点,定义了连续属性样本分类概念和分割点评价指标,提出了一种新的连续属性样本分类规则挖掘算法.选择一个试验区,采用该算法分别对遥感光谱数据、遥感光谱和DEM数据相结合的数据进行分类规则挖掘、遥感影像分类和分类精度比较.结果表明：（1）该算法具有较高的分类精度;（2）加入DEM等与分类相关的其他空间数据可以提高遥感影像的分类精度.通过挖掘分类规则进行遥感影像分类,扩展了基于知识的逻辑推理分类方法中知识获取渠道,提高了分类规则获取的智能化程度.新的连续属性样本分类规则挖掘算法,扩展了归纳学习算法对连续属性样本分类的适应性.     

Fuzzy Classification for Mapping Invasive Species from Multispectral Imagery

Vegetation maps are essential tools for planning and evaluation in nature conservation. To a large extend, management objectives can be defined in terms of vegetation attributes. Vegetation maps are frequently used in conservation planning and evaluation. Nowadays, remote sensing is one of the most important sources of vegetation cover classifications at different scales. In this paper, the usefulness of moderate resolution images for invasive species Cirsium arvense and Stachys byzanthina mapping at local scale is evaluated. A fuzzy classification approach is tested for improved discrimination of invasive species. The study was carried out in Vazroud rangelands located in the Mazandaran province of Iran. In this study among the analysed Z-score for uncertainty map creation is found that 0.9, 1.0 for TM image and 1.1 for IRS image in normalized are the best based on known attributes of the study area.