Accuracy test of point-based and object-based urban building feature classification and extraction applying airborne LiDAR data

Point-based and object-based building extractions were conducted in airborne LiDAR data in a sample area of Buffalo, New York. First, the earth surface points were filtered from the entire laser scan data set using a new filtering algorithm, which combines the TIN slope modelling and statistical analysis. The off-ground points were extracted for buildings in the study area using both point cluster analysis and object-oriented classifications. The accuracies of both approaches were tested using the digitised ground truth. The outcomes of accuracy testing of the point-based method are correctness: 88.74%, completeness: 92.67% and quality: 83.50%. The results of the accuracy of object-based building extraction are correctness: 87.21%, completeness: 60.14%, and quality: 55.26%. Reconstructions of 3D building models based on the extracted building points were performed. This study contributes scientific and technological knowledge for researchers in developing more effective methods in converting the LiDAR survey to a 3D GIS database.     

基于面向对象的高分辨率影像专题信息提取

利用面向对象的分类方法对高分辨率影像上的水体、植被、建筑物以及道路进行了自动提取;对自动提取结果进行了精度评价:水体和植被的用户精度在90%以上,建筑物和道路中心线的用户精度在70%以上;并分析了影响信息提取精度的因素。     

基于易康平台的高分辨率影像在测绘生产中的地物提取

随着遥感技术的发展,遥感数据获取手段有了长足的进步。测绘部门在生产中也展开了基于高分辨率影像提取地表覆盖等信息。本论文以易康软件作为卫星遥感影像地物提取的平台,研究如何利用高分辨率遥感影像生成地表覆盖分类图,以达到测绘部门在生产中的需求。     

基于网络微博的地震宏观异常信息提取研究——以芦山地震为例

微博平台有用户群大、公众参与性强、实时性等优点,同时微博平台信息又具有信息真伪难辨、地址信息模糊等缺点.本文以芦山地震为例,针对微博内容如何提取和地址如何定位两方面进行了分析研究,对于如何在网络微博平台中及时的提取地震宏观异常信息,提出了聚焦爬虫技术,并对微博地址进行了分类,同时将正向最大匹配和特征词地址分词的中文地址匹配模型应用于地址信息的提取和地址匹配中;最后将不同的地址类别定位为不同的行政级别,使微博平台和微博信息得到了充分的利用.通过研究认识到微博信息在反应震前异常的发生趋势方面有一定的参考价值（动物异常和气象异常所占比例较大）,是不能被忽略的：地址方面可以看出异常随着时间的逼近有向震中聚集的趋势,有一定的参考价值.     

Areal interpolation of spatially extensive variables: A comparison of alternative techniques

Abstract

This paper presents a methodology which provides a practicable solution to a specific class of areal interpolation problems. The method allows for the transformation of the areal basis of land use data from a set of reporting units to a set of land capability classes within each of which land use mix is assumed to be homogeneous. Three alternative procedures, based on least squares criteria, are suggested for estimation of the land use mix in each land class. Although these are shown to produce similar and broadly plausible results in an empirical application, choice of procedure may be guided by consideration of the particular application and computational ease.     

利用最小交叉熵的多时相影像变化检测

针对目前多时相遥感影像变化检测中缺少有效的确定变化阈值方法的问题,文章提出一种基于最小交叉熵的方法:采用中值滤波方法分别消除两个时相遥感影像中的噪声;将比值法和差值法融合构造两个时相遥感影像的差异影像;通过最小交叉熵确定差异影像的最佳变化阈值,并对差异影像进行分割得到变化区域;最后对变化检测精度进行评估。实验结果表明,提出的方法具有较高的检测精度,是一种有效的多时相遥感影像变化检测方法。     

多尺度多准则的遥感影像线状地物信息提取

针对传统低层次单个像元纯光谱的信息提取方法存在的不足,以ZY-3高分辨率遥感影像为数据源,依据"种内一致性最大、类间异质性最大"原则,提出了基于多尺度多准则的ZY-3影像线状地物信息提取方法。利用基于最优分割尺度的影像分割、面向对象线状地物目标信息的提取方法,实现了河流、水渠、道路等线状地物信息的提取,并通过实验分析验证了可行性、有效性。结果表明:构建的RMAS指数指标,可以作为多尺度分割中选择最优分割尺度的一种很好的指示器,所提方法不仅有效避免了分割斑块面积对分割结果评价的影响,还能够有效识别ZY-3影像中的线状地物信息,研究结果对高分辨率遥感影像处理尺度依赖的技术难点及线状地物信息提取等研究具有重要的参考价值。     

运动分割和光流估计的卫星视频超分辨率重建

针对卫星视频中存在局部运动目标的问题,该文提出了一种运动分割和光流估计联合约束的卫星视频超分辨率重建算法。首先通过视频帧间运动分割方法提取局部运动目标,获得其边缘信息;其次用光流方法对视频图像进行逐像素的运动估计,获得初始运动矢量场;利用分割边缘和运动矢量场进行联合约束,并统一运动目标内的光流矢量,获得纠正后的运动矢量场,并作为几何运动矩阵,将其补偿到超分辨率重建模型中生成高分辨率图像;最后用卫星视频数据进行实验,并与传统方法结果进行对比,证明该文算法可以有效地克服运动目标的伪影效应,提升了卫星视频的分辨率。     

An improved tree crown delineation method based on live crown ratios from airborne LiDAR data

Individual tree crowns are one of the basic forest inventory data, which can be used in various forest-related studies such as biomass and carbon stock estimation. High-resolution remote-sensing data including airborne LiDAR-derived surfaces have been widely used for delineating tree crowns. This study proposes an improved tree crown delineation algorithm that can be effectively applied to a range of forests with a limited number of parameters considering its operational use with airborne LiDAR data. The proposed algorithm integrates morphological operators, Otsu’s method, marker-controlled watershed segmentation, and the concept of crown ratios. The proposed algorithm was compared with the region growing method, a widely used tree crown delineation algorithm. The two algorithms were evaluated over 10 plots in rugged terrain located in Kangwon Province in South Korea. Results show that the proposed approach produced much better performance (~87% matched on average) for 10 plots with a range of tree densities than the region growing method (~60% matched on average). The proposed algorithm worked better for sparse plots than dense ones. It also worked well for deciduous plots (plots 1 and 4). On the other hand, the region growing method produced relatively low accuracy with many merged crowns, which requires additional postprocessing such as a resplit step.     

面向小间距地图注记的智能识别研究

在充分分析地图注记特点的基础上,设计针对小间距地图注记的智能识别系统,提出分析投影积分高坡值个数的文本图像倾斜校正方法,并且,基于地图注记为印刷稿字体,其相同字号字体占位一致的特点,提出基于间距积分分割文本行图像的方法,最后利用现有小型OCR系统常用的基于结构化和统计信息相结合的特征提取方法构建特征字库,基于此特征数据库对提取的地图注记字符进行识别。