基于ASTER数据遥感影像的决策树分类

以黑龙江省北安市为研究区域,尝试利用ASTER视反射率值进行便利、准确的土地利用分类研究。对ASTER数据进行波段相关分析,确定最佳组合波段;然后重点分析转换为视反射率值的影像特征和光谱特征,从中提取各种典型地物的光谱曲线; 并依据提取的光谱曲线建立基于地物反射率值大小关系或阈值的决策树模型,对研究区不同地物类型进行分类,并对结果进行精度评价。应用效果表明,该方法简单有效,但对于混合光谱容易错分。     

基于决策树模型的上海城市湿地遥感提取与分类

城市湿地是上海重要的生态基础并具有复杂多变的自然特性。研究采用决策树分类方法,以TM影像多光谱波段特征为主要分类变量,采用经K-T变换、IHS变换等光谱增强后的数据以及利用灰度共生矩阵分析影像第一主成分的纹理统计量作为辅助分类变量,结合城市湿地几何特征信息,构建上海城市湿地决策树分类模型,进行上海市湿地信息的遥感提取和分类。结果表明：（1）上海城市湿地总面积为1 277.40 km²;其中水田面积最大,占总面积的65.30%;其次为河流、库塘、湖泊和芦苇。（2）决策树模型的分类方法在一定程度上提高了城市湿地提取和分类的精度,使其达到89.05%;与传统的最大似然法相比,总精度提高了约10%。     

ASTER数据在深圳东部滨海地区土地利用调查中的应用

以深圳市东部滨海地区为试验区,对2004年11月21日ASTER遥感数据进行辐射和几何精校正处理,实地建立分类样地;根据多边形样地矢量数据计算分析12类地物在ASTER各波段光谱反射图和分类叠合图,同时进行植被指数和短波红外5个波段主成分分析;结合GIS并利用ASTER光谱波段、第一主成分、植被指数、立体像对生成的地形因子建立土地利用分类决策树表;再根据决策树表对ASTER影像进行土地利用分类。经验证,分类结果总体精度达到85.1%。应用效果表明,利用ASTER数据进行土地现状资源调查具有很好的性价比,能够满足土地利用现状调查的准确度和精度。     

利用粗糙集理论进行遥感分类信息提取

以辽宁省双台子河口湿地为研究对象,以Landsat 8和HJ-1-A/HJ-1-B的多时相遥感影像为数据源,根据研究区现状,将研究区分为旱地、芦苇、水田、碱蓬、混合植被、水面、滩涂、居民点、养殖塘九个类型.利用时间序列的归一化植被指数提取植被与非植被的分类阈值,采用粗糙集理论和多时相遥感影像,对植被和非植被分别进行分类规则的获取,建立了研究区决策树分类模型.为了进行精度评价,利用相同的训练点又进行了同样基于像元的最大似然法分类.最后利用混淆矩阵对上述两种方法进行了精度评估,基于粗糙集的决策树分类法与最大似然法总体分类精度分别为93.70%和91.62%,Kappa系数分别为0.92和0.90,两项指标值基于粗糙集理论法均比最大似然法有所提高.这为构建决策树分类模型进行湿地地表分类信息提取提供了一条新的研究思路.     

三峡库区岩性智能分类研究

三峡库区岩体上方覆盖着厚实的土壤和茂密的植被,是高植被覆盖区,岩性信息弱,因此岩性识别和分类困难,没有成熟的方法可循。针对三峡库区进行岩性分析,选择三峡库区巴东城区作为研究区域,采用2000年5月成像的ETM+遥感影像,构造纹理、光谱、植被覆盖等17个分类因子,将遥感影像与地质图叠加,选取1 101个样本点,采用决策树C4.5算法,挖掘出三峡库区巴东县处岩性的解译规则和知识,决策树的学习精度为96.6%,剪枝后精度为95.9%,规则提取的精度为93.1%,提取的规则置信度很高,并基于知识驱动和规则匹配实现了岩性的智能分类,分类精度较高为90.11%;将分类结果与IsoData方法、K-Means方法、马氏距离法、最大似然法、最小距离法、平行六面体方法等6种方法的分类结果进行比较,试验结果证明,决策树方法的分类结果最好,精度明显高于其他6种方法。     

面向对象的喀斯特地区土地利用遥感分类信息提取

传统的面向像元分类方法虽然对光谱差异较为明显的遥感影像信息提取具有较好的效果,但会不可避免地产生“椒盐现象”,同时对纹理和形状信息不能充分应用,造成了大量信息损失。为了提高喀斯特地区土地利用遥感信息提取的精度,本文采用面向对象的分类方法,对贵州省毕节地区开展了土地利用遥感信息自动提取研究。首先对该地区Landsat-5 TM影像进行多尺度分割,形成影像对象层,然后综合应用基于知识决策树分类和基于样本的最邻近分类等技术对喀斯特地区进行遥感解译。结果表明,面向对象分类技术能较好地对喀斯特地区土地利用信息进行提取,同时避免了“椒盐现象”的产生,经野外采集样点数据验证,一级类分类精度为91.7 %,二级类分类精度为89.4 %,表明该方法在贵州省毕节地区应用效果良好。      

基于MODIS数据的朝鲜半岛土地利用变化信息提取

以朝鲜半岛为研究区域, 基于2000年和2007年的MOD IS数据, 应用线性光谱混合模型进行像元组分分解并提取分类特征, 对像元组分分解后有错分的地物类型结合纹理进行分析, 以决策树模型进行土地利用分类。结果表明, 像元组分分解后的决策树分类结果总体精度达78.346 5% , Kappa系数达0.681 3。与像元组分分解前最大似然法和决策树分类结果相比, 决策树分类精度优于最大似然法, 且像元组分分解方法提高了分类精度。经2000年和2007年像元组分分解后的决策树分类结果对比: 朝鲜耕地面积增加; 韩国耕地面积减少, 居民地面积增加。     

基于时序MODIS NDVI的黑河流域土地覆盖分类研究

归一化植被指数(NDVI)是植被生长状态及植被覆盖度的最佳指示因子,其时序数据也已成为基于生物气候特征开展大区域植被和土地覆盖分类的基本手段。基于时序NDVI数据的土地覆盖分类,即通过提取NDVI时间信号所包含的植被生物学参数,构建起一个包含植被生物学信息的分类特征空间。利用2006年重建得到的MODIS NDVI 16天合成时间序列数据,并结合1 km分辨率的DEM数据、野外实地调查资料等辅助数据,综合分析了不同土地覆盖类型对应的时序NDVI谱线及其第一、二谐波的特征阈值,建立决策树对黑河流域的土地覆盖开展分类研究。结果表明,基于时序MODIS NDVI谱线特征的决策树分类精度为78%,Kappa系数为0.74。利用1 km时序MODIS NDVI时间序列获得较为准确的黑河流域土地覆盖类型是可行的。     

机载LiDAR数据加工DEM产品技术研究

重点讨论基于不同软件平台下,机载LiDAR数据加工DEM测绘产品工艺技术研究。针对原始LiDAR点云集数据拼接、系统精度纠正、预处理→地面高程点提取→非地面点滤波分类→数字地面模型编辑→数字高程模型网格化→DEM产品质量检验等生产工艺流程进行研究。通过对LiDAR数据滤波分类处理算法应用研究,实现不同地物类别点云自动化提取。同时针对高精度的DTM、DEM数模产品拓展应用,重构地表模型三维虚拟成像。     

决策树方法在遥感地质填图中的应用

决策树理论在遥感分类中, 分类准确、高效.依据其理论方法, 对青海省民和地区的遥感数据———ETM + (enhanced thematic mapper plus) 进行了分类, 选用的ETM +数据为1999年10月份数据, 数字高程(DEM) 数据来自于1:2 5万民和幅地形图, 数据格式为MapInfo通用格式MIF, 数据进行了坐标转换(地理坐标), 对原始数据进行了处理, 从等高线中提取数字高程.对遥感数据进行地形及光照矫正, 计算植被因子及缨帽变换的3个分量, 同其他5个遥感波段结合形成原始分类图层, 同时确定目标分类结果.原始数据的采样基于目视, 首先采用不同的彩色合成方案突出不同的目标地物, 交互式进行采样, 使用IDL语言编制程序从原始数据中提取地物数字信息, 使用Clementine7.2对数据进行处理, 其中10 %的采样数据验证模型准确率, 其余数据用来推算模型, 对数据进行10次迭代, 同时给予75 %的剪枝, 得到区分不同地物(如红层、黄土等) 的最合适图层(band 1 & band 3)和具体数值, 形成决策树模型, 将决策树模型导入Envi4.0中, 对原始数据(9个图层) 进行计算形成初步分类结果图, 对初步分类结果图进行一定的碎片合并, 最终形成分类结果图.该图同1:2 5万地质图进行对比确认分类的效果, 同传统分类图比较确认决策树分类方法优于传统分类.另外来自于决策树所提取的信息, 有利于地学知识的归纳总结      

露天灰岩矿纹理特征分析及面向对象的分类

矿产资源过度开发造成了严重的环境污染,可持续发展面临严峻挑战,矿山治理迫在眉睫。遥感信息提取技术能够便捷地获取矿山开发信息,为矿山治理提供依据。然而,目前信息提取的精度和自动化程度仍有提升空间,并且就矿产类型而言,关于石灰岩矿开采信息提取的研究甚少。针对以上现状,本文以唐山市的一处灰岩矿山为例,以Landsat-8影像作为基础数据,通过计算不同地物的纹理信息,包括地物均值、方差、信息熵、偏离度和数据范围,构造了能够突出露天灰岩矿开采区域的纹理指数模型(texture variance index,简称TVI)。结合已有的光谱和空间特征,设定准确的阈值并建立决策树,以面向对象方法为依托,进行露天灰岩矿山开采边界的信息提取。结果表明,将纹理方差指数作为灰岩矿山开发信息提取的规则之一,总体精度、用户精度和Kappa系数与原有决策树相比皆有所提高。以文中灰岩矿开采范围提取结果为基础,根据高分辨率影像的纹理特征,进一步实现了新矿区与待复垦旧矿区的分类,总体精度约为0.916。     

Sugarcane crop identification from LISS IV data using ISODATA,MLC, and indices based decision tree approach

Image classification is one of the crucial techniques in detecting the crops from remotely sensed data. Crop identification and discrimination provide an important basis for many agricultural applications with various purposes, such as cropping pattern analysis, acreage estimation, and yield estimation. Accurate and faster estimation of crop area is very essential for projecting yearly agriculture production for deciding agriculture policies. Remote sensing is a technique that allows mapping of large areas in a fast and economical way. In many applications of remote sensing, a user is often interested in identifying the specific crop only while other classes may be of no interest. Indian Remote Sensing Satellite (IRS-P6) LISS IV sensor image of spatial resolution 5.8 m has been used to identify the sugarcane crop for the Chhapar village of Muzaffarnagar District, India. Classification of satellite data is one of the primary steps for information extraction for crop land identification. In recent years, decision tree approach to image analysis has been developed for the assessment and improvement of traditional statistically based image classification. In this study, ISODATA, MLC, and vegetation indices based decision tree approaches are used for classifying LISS IV imagery. The 11 vegetation index images have been generated for decision tree classification. All the three methods are compared and it is found that the best performance is given by the decision tree method. Vegetation indices based decision tree method for sugarcane classification, the user’s accuracy, producer’s accuracy, overall accuracy, and kappa coefficient were found 88.17, 86.59, and 87.93% and 0.86 respectively.     

Accuracy Analysis of Digital Elevation Model Relating to Spatial Resolution and Terrain Slope by Bilinear Interpolation

This study investigates the accuracy analysis of the digital elevation model (DEM) with respect to the following two major factors that strongly affect the interpolated accuracy: (1) spatial resolution of a DEM and (2) terrain slope. Unlike existing studies based mainly on a simulation approach, this research first provides an analytical approach in order to build the relationship between the interpolated DEM accuracy and its influencing factors. The bi-linear interpolation model was adopted to produce this analytic model formalized as inequalities. Then, our analytic models were verified and further rectified by means of experimental studies in order to derive a practical formula for estimating the DEM accuracy together with an optimization model for calculating the required resolution when a prescribed upper bound to the DEM accuracy is given. Moreover, this analytic approach can cope with either a grid-based DEM or a randomly scattered scenario whose efficacies have been validated by the experiments using both synthetic and realistic data sets. In particular, these findings first establish the rules for directly correlating the horizontal resolution of DEM data with vertical accuracy.     

The assessment of landslide susceptibility mapping using random forest and decision tree methods in the Three Gorges Reservoir area,China

Landslide susceptibility mapping is an indispensable prerequisite for landslide prevention and reduction. At present, research into landslide susceptibility mapping has begun to combine machine learning with remote sensing and geographic information system (GIS) techniques. The random forest model is a new integrated classification method, but its application to landslide susceptibility mapping remains limited. Landslides represent a serious threat to the lives and property of people living in the Zigui–Badong area in the Three Gorges region of China, as well as to the operation of the Three Gorges Reservoir. However, the geological structure of this region is complex, involving steep mountains and deep valleys. The purpose of the current study is to produce a landslide susceptibility map of the Zigui–Badong area using a random forest model, multisource data, GIS, and remote sensing data. In total, 300 pre-existing landslide locations were obtained from a landslide inventory map. These landslides were identified using visual interpretation of high-resolution remote sensing images, topographic and geologic data, and extensive field surveys. The occurrence of landslides is closely related to a series of environmental parameters. Topographic, geologic, Landsat-8 image, raining data, and seismic data were used as the primary data sources to extract the geo-environmental factors influencing landslides. Thirty-four layers of causative factors were prepared as predictor variables, which can mainly be categorized as topographic, geological, hydrological, land cover, and environmental trigger parameters. The random forest method is an ensemble classification technique that extends diversity among the classification trees by resampling the data with replacement and randomly changing the predictive variable sets during the different tree induction processes. A random forest model was adopted to calculate the quantitative relationships between the landslide-conditioning factors and the landslide inventory map and then generate a landslide susceptibility map. The analytical results were compared with known landslide locations in terms of area under the receiver operating characteristic curve. The random forest model has an area ratio of 86.10%. In contrast to the random forest (whole factors, WF), random forest (12 major factors, 12F), decision tree (WF), decision tree (12F), the final result shows that random forest (12F) has a higher prediction accuracy. Meanwhile, the random forest models have higher prediction accuracy than the decision tree model. Subsequently, the landslide susceptibility map was classified into five classes (very low, low, moderate, high, and very high). The results demonstrate that the random forest model achieved a reasonable accuracy in landslide susceptibility mapping. The landslide hazard zone information will be useful for general development planning and landslide risk management.     

A method for mineral prospectivity mapping integrating C4.5 decision tree,weights-of-evidence and m-branch smoothing techniques: a case study in the eastern Kunlun Mountains,China

In this study, a novel method that integrates C4.5 decision tree, weights-of-evidence and m-branch smoothing techniques was proposed for mineral prospectivity mapping. First, a weights-of-evidence model was used to rank the importance of each evidential map and determine the optimal buffer distance. Second, a classification technique that uses a C4.5 decision tree in data mining was used to construct a decision tree classifier for the grid dataset. Finally, an m-branch smoothing technique was used as a predictor, which transformed the decision tree into a probability evaluation tree. The method makes no conditional independence assumption and can be applied for class imbalanced datasets like those collected during mineral exploration for prospectivity mapping of an area. The traits of comprehensibility, accuracy and efficiency were derived from the C4.5 decision tree. In addition, a case study for iron prospectivity mapping was performed in the eastern Kunlun Mountains, China. Sixty-two Skarn iron deposits and eight evidential maps related to iron mineralization were studied. On the final map, areas of low, moderate and high potential for iron deposit occurrence covered areas of 71,491, 14,298, and 9,532 km², respectively. For the goodness-of-fit test, 91.94 % of the total 62 iron deposits were within a high-potential area, 8.06 % were within a moderate-potential area and 0 % were within a low-potential area. For ten-fold cross-validation, 82.26 % were within a high-potential area, 14.52 % were within a moderate-potential area and 3.22 % were within a low-potential area. To evaluate the predictive accuracy, Receiver Operating Characteristic (ROC) curves and Area Under ROC Curve (AUC) were employed. The accuracy of the goodness-of-fit test reached 97.07 %, and the accuracy of the ten-fold cross-validation was 95.10 %. The majority of the iron deposits were within high-potential and moderate-potential areas, which covered a small proportion of the study area.     

区域地球化学元素分布与组合求异方法

在与单元素异常处理方法比较基础上,应用多元素空间分布与空间定量组合求异方法,以吉林白山地区1∶ 20 万化探测量数据处理分析为例,进行空间定量组合求异模型研究,建立了该区的中大比例尺因子--泛克里格模型,应用DEM 原理表达三维元素空间分布模型,并给出地质解译。结果表明: 多元空间定量组合求异法较传统单元素求异法不仅精度更高,更能反应元素空间的内在结构与关联性,且提供了更多的地学信息,尤其使DEM 的三维表达更为直观。