电子地图中线状要素移动注记关键问题研究   总被引：4，自引：0，他引：4  

廖楚江  蔡忠亮  杜清运  姜坤丽 《测绘信息与工程》2004,29(4):26-28

探讨了电子地图移动注记的几个关键问题 ,提出了将分段注记和移动注记相结合的思想 ,并在此基础上提出了采用回溯法解决注记之间的冲突 ,从而较好地解决了图幅移动之后某些屏幕内线状要素注记不可见或表达模糊的问题 ,实现了注记输出的完整性和美观性的统一 ,提高了图形输出的效率 ,节省了存储空间 ,在实际应用中取得了较好的效果 ,充分体现了电子地图在线状要素注记显示这一环节的优越性  相似文献   

集成高度和彩色纹理特征的影像目标模糊聚类识别方法     

潘励  郑宏  张祖勋  张剑清 《武汉大学学报(信息科学版)》2004,29(4):311-314

提出一种基于高度和彩色纹理信息的目标识别方法 ,其目的是提取具有相对高度的地物。由于传统二维纹理分析的方法往往不能有效地提取这类目标 ,本文把高度、色彩和纹理信息有机地结合起来 ,提高了识别率。依据影像提取结果 ,探讨了树林对森林覆盖地区自动空中三角测量选点的负面影响 ,同时建议采用数据分组的方案来改善森林覆盖地区自动空中三角测量的精度 ,并用对比实验证明了其有效性。  相似文献   

一种组合优化的多边形化简方法   总被引：2，自引：0，他引：2  

杜维  艾廷华  徐峥 《武汉大学学报(信息科学版)》2004,29(6):548-550

以多边形轮廓为目标 ,依据曲线特征点将其分解为一系列的弯曲特征 ,并对此弯曲特征集实施组合优化 ,将入围弯曲首尾相连 ,即可得到最终的化简结果  相似文献   

GEOCHEMICAL CHARACTERISTICS OF PLATINUM GROUP ELEMENTS IN JINCHUAN SUPER-LARGE SULFIDE COPPER-NICKEL DEPOSIT, JINCHANG CITY, GANSU PROVINCE, CHINA     

WANG Ruiting MAO Jingwen HE Ying WANG Dongsheng TANG Zhongli 《大地构造与成矿学(英文版)》2005,29(2):152-163

The platinum-group element geochemistry of rocks and ores from Jinchuan super-large copper-nickel sulfide deposit is systemically studied in this paper. The Cu/Pd mean ratio of Jinchuan intrusion is lower than that of original mantle magma, which indicates that these ultrabasic rocks were crystallized from magma that lost Pd in the form of melting segregation of sulfides. The PGE of the rocks show trend of partial melting, similar to that of mantle peridotite, which shows that magma formation occurs during rock-forming and ore-forming processes. The chondrite normalized PGE patterns of the rocks and ores are well related to each other, which signifies the signatures of multi-episode magmatic intrusion, melting and differentiation in the formation processes of rocks and ores. In addition, analyses about the relation between PGE and S, and study on Re-Os isotopes indicate that few contamination of the crustal substances occurred during the magmatic intrusion and the formation of deposit. However, contamination by crustal substances helps to supply part of the S for the enrichment of PGE. Meanwhile, the hydrothermal process is also advantageous for the enrichment of PGE, especially lbr Pt and Pd, due to deep melting segregation. The characteristic parameters （such as Pt/（Pt＋Pd）, （Pt＋Pd）/（Ru＋Ir＋Os）, Pd/Ir, Cu/（Ni＋Cu）, and so on.） for platinum-group elements for Jinchuan sulfide copper-nickel deposit show the same features as those for sulfide copper-nickel deposit related to basic magma, which also illustrates its original magma property representative of Mg-high tholeiite. Therefore, it is the marie （not ultramafic） magma that resulted in the formation of the superlarge sulfide copper-nickel deposit enriched in Cu and PGE. To sum up, the geochemical characteristics of platinum-group elements in rocks and ores from Jinchuan copper-nickel sulfide deposit are constrained by the continental rift tectonic environment, the parent magma features, the enriched mantel magma source, the complex metallogenesis and PGE geochemical signatures, and this would be rather significant for the study about the genetic mechanism of copper-nickel sulfide deposits.  相似文献   

基于随机森林特征选择的城市绿化乔木树种分类     

温小乐  钟奥  胡秀娟 《地球信息科学学报》2018,20(12):1777-1786

城市绿化在改善空气、水和土壤质量,吸收和减少二氧化碳及各种污染物,缓解城市热岛和减少雨水径流等方面发挥着重要作用。及时准确地获取树种信息是城市规划与绿化管理的先决条件,对进一步改善城市生态环境也具有重要意义。基于遥感技术,使用高空间分辨率的WorldView-2卫星影像,采用光谱、纹理、指数以及几何等多种特征相结合的面向对象方法,并通过随机森林进行特征选择,对福州大学旗山校区北部的榕树、杧果、香樟、重阳木、羊蹄甲、垂叶榕以及木棉7种主要绿化乔木进行树种分类。实地验证结果表明：通过特征选择可以减少或规避数据冗余以及休斯效应的产生,该方法可以提高现有同类型树种分类的精度,当淘汰全部特征的20%,利用34个特征（包括15个光谱特征、6个纹理特征、8个指数特征和5个几何特征）进行分类时,总精度最高,可达74.95%,Kappa系数为0.67。其中,光谱平均值的特征重要性最高,而各波段的标准差的重要性较低。WorldView-2卫星影像的4个新增波段,特别是黄光和红边波段及其构建的指数特征重要性较高,也说明这些波段在植被遥感,特别是树种分类中极具应用前景。  相似文献   

一种多层特征融合的道路场景实时分割方法     

周继苗  李必军  陈世增 《测绘通报》2020,(1):10-15

道路场景理解是自动驾驶领域中重要模块之一，它可以提供关于道路更丰富的信息，对于建立高精度地图和实时规划都具有重要作用。其中，语义分割可以为图像每个像素赋予类别信息，是自动驾驶场景理解中最常用的方法。但是，目前常用的语义分割算法在速度和精度上大都不能达到很好的平衡。本文在MobileNetV2的基础上，提出了一种多层次特征融合的方法，使得网络可以在实时运行的同时保证精度满足实际应用的需求，并在Cityscapes数据集上进行了试验验证和分析。  相似文献   

顾及局部与结构特征的稀疏多项式逻辑回归高光谱图像分类方法     

沈宇臻  官云兰  杨禄  刘承承  严小芳 《测绘通报》2019,(6):24-28

稀疏多项式逻辑回归在分类中仅利用图像光谱信息，导致分类效果不太理想。本文提出了一种顾及局部与结构特征的稀疏多项式逻辑回归高光谱图像分类方法。首先利用加权均值滤波与拓展形态学多属性剖面对原始高光谱图像进行局部与结构特征提取；然后对二者进行加权平均特征级融合以获取更具唯一性的像元特征；最后由稀疏多项式逻辑回归分类器对融合结果进行分类。结果表明，本文方法能有效地提高分类精度，而且具有较强的稳健性。  相似文献   

基于深度残差网络的机载LiDAR点云分类     

赵传  郭海涛  卢俊  余东行  张保明 《测绘学报》2020,49(2):202-213

机载LiDAR点云的分类是利用其进行城市场景三维重建的关键步骤之一。为充分利用现有的图像领域性能较好的深度学习网络模型,提高点云分类精度,并降低训练时间和对训练样本数量的要求,本文提出一种基于深度残差网络的机载LiDAR点云分类方法。首先提取归一化高程、表面变化率、强度和归一化植被指数4种具有较高区分度的点云低层次特征;然后通过设置不同的邻域大小和视角,利用所提出的点云特征图生成策略,得到多尺度和多视角点云特征图;再将点云特征图输入到预训练的深度残差网络,提取多尺度和多视角深层次特征;最后构建并训练神经网络分类器,利用训练的模型对待分类点云进行预测,经后处理得到分类结果。利用ISPRS三维语义标记竞赛的公开标准数据集进行试验,结果表明,本文方法可有效区分建筑物、地面、车辆等8类地物,分类结果的总体精度为87.1%,可为城市场景三维重建提供可靠的信息。  相似文献   

多尺度邻域特征下的机载LiDAR点云电力线分类   总被引：1，自引：0，他引：1  

王艳军  李凯  路立娟 《测绘通报》2019,(4):21-25

利用机载激光雷达技术三维测量精度高且获取快速的优点进行电力线自动分类提取已成为点云数据处理与电力应用的重要领域。针对电力线分类模型的自动化和高精度需求，本文提出了基于三维多尺度邻域特征的机载LiDAR点云电力线分类提取模型框架，主要包括4个步骤：电力线候选点滤波、多尺度邻域类型选取、形状结构特征提取和支持向量机分类。通过对2个复杂城市区域的试验数据集和8种不同邻域类型的详细结果对比分析，发现基于多尺度圆球邻域形状结构特征的分类模型结果准确率、召回率和质量分别达到97%、94%和93%，同时整体处理时间在2个试验数据中分别从366、256 s减少到274、160 s。试验结果表明，该方法在多种复杂城市场景中能够实现机载LiDAR数据的电力线较高精度分类提取。  相似文献   

Watershed delineation using hydrographic features and a DEM in plain river network region          下载免费PDF全文

Zhengqing Lai  Shuo Li  Guonian Lv  Zhirong Pan  Guosong Fei 《水文研究》2016,30(2):276-288

Watershed delineation is a required step when conducting any spatially distributed hydrological modelling. Automated approaches are often proposed to delineate a watershed based on a river network extracted from the digital elevation model (DEM) using the deterministic eight‐neighbour (D8) method. However, a realistic river network cannot be derived from conventional DEM processing methods for a large flat area with a complex network of rivers, lakes, reservoirs, and polders, referred to as a plain river network region (PRNR). In this study, a new approach, which uses both hydrographic features and DEM, has been developed to address the problems of watershed delineation in PRNR. It extracts the river nodes and determines the flow directions of the river network based on a vector‐based hydrographic feature data model. The river network, lakes, reservoirs, and polders are then used to modify the flow directions of grid cells determined by D8 approach. The watershed is eventually delineated into four types of catchments including lakes, reservoirs, polders, and overland catchments based on the flow direction matrix and the location of river nodes. Multiple flow directions of grid cells are represented using a multi‐direction encoding method, and multiple outflows of catchments are also reflected in the topology of catchments. The proposed approach is applied to the western Taihu watershed in China. Comparisons between the results obtained from the D8 approach, the ‘stream burning’ approach, and those from the proposed approach clearly demonstrate an improvement of the new approach over the conventional approaches. This approach will benefit the development of distributed hydrological models in PRNR for the consideration of different types and multiple inlets and outlets of catchments. Copyright © 2015 John Wiley & Sons, Ltd.  相似文献