基于车载点云数据的树木提取与分析

本文基于高速公路高精度点云数据,首先通过点云数据的分类处理实现对树木点云数据的提取,将树木点云投影到水平面,采用DBSCAN密度聚类算法实现单根树木的提取;然后在数据密集区域存在树木树冠点云重叠的区域,本文结合树干几何特征提取树干的位置信息,计算所有点云到树干中心的欧氏距离,将所有点云归类到最近的树干进行粗分割;最后根据粗分割的树木轮廓特征确定树冠模型与树冠中心,提出了采用基于密度特征的格网竞争算法对重叠的区域进行精细分割。试验表明,本文采用的树木分割方法能够实现单棵树木精确提取。     

交通监控视频图像语义分割及其拼接方法

视频拼接是图像拼接的外延,在场景监控、目标识别等应用中发挥着重要作用。传统视频拼接算法多要求视频间具有较大重叠区域且特征点匹配过程中只顾及图像几何特征,当处理交通监控视频时,会因不同摄像头之间重叠区域极小或主光轴之间夹角较大而导致无法拼接或图像变形较大。为此,本文提出一种交通监控视频图像语义分割及其拼接方法。首先,利用边缘角度二阶差分直方图算法自动识别多视频交汇区域的正射影像,并将其作为拼接背景图像;然后,基于全卷积神经网络对正射影像和视频图像进行语义分割,提取图像中的交通专题语义;最后,以交通专题语义作为约束进行特征点匹配,将各个交通监控视频匹配至背景正射影像,实现监控区域视频拼接。采用山东省某市实际视频数据进行试验验证,结果表明对于重叠区域较小的监控视频,本文方法可获得较好地拼接图像,同时可有效提高特征点匹配的准确度。     

条件随机场模型约束下的遥感影像模糊C-均值聚类算法

遥感影像具有丰富的空间相关信息,而传统的基于像元光谱的聚类算法并不能将空间信息融入聚类,聚类结果往往不好。针对这一问题,本文提出了一种条件随机场模型约束下的模糊C-均值聚类算法,通过邻域像元的分类先验信息对中心像元的类别进行约束从而提取空间相关信息,基于二阶条件随机场将光谱信息和空间相关信息同时融入聚类,并使用环形置信度迭代算法得到像元分类后验概率的全局最优推测。试验证明,本文算法能够有效地保持地物的形状特征,分类精度相比传统算法有所提高。     

Spatial and temporal variation of global LAI during 1981–2006

Earth is always changing. Knowledge about where changes happened is the first step for us to understand how these changes affect our lives. In this paper, we use a long-term leaf area index data (LAI) to identify where changes happened and where has experienced the strongest change around the globe during 1981–2006. Results show that, over the past 26 years, LAI has generally increased at a rate of 0.0013 per year around the globe. The strongest increasing trend is around 0.0032 per year in the middle and northern high latitudes (north of 30°N). LAI has prominently increased in Europe, Siberia, Indian Peninsula, America and south Canada, South region of Sahara, southwest corner of Australia and Kgalagadi Basin; while noticeably decreased in Southeast Asia, southeastern China, central Africa, central and southern South America and arctic areas in North America.     

面向道路网的GPS浮动车覆盖率模型研究

给出了对ISO19115-3进一步扩展的方案,完善了影像与栅格元数据模型。利用UML概念与OWL概念之间的映射关系,实现了基于OWL表达的影像与栅格元数据规范本体,并利用该本体描述提出了一种基于用户交互的语义扩展方式,并将其应用到目录服务元数据查询中,缩小了用户查询与元数据描述的语义鸿沟,提高了目录服务查询的针对性,并为影像与栅格元数据的发现提供了一种更加人性化的交互方法。     

基于时空模糊本体的交通领域知识建模

在分析本体技术和交通信息特点的基础上,针对交通概念描述的特殊要求,提出了时空模糊本体的概念,以描述交通领域概念,并利用时空模糊本体进行交通领域知识建模。     

利用自组织网络分析1990～2010年中国主要城市扩展特征

以我国27个主要城市为研究样木,采用遥感数据和统计数据获取1990-2010年间的城市扩展和社会经济信息,从城市扩张速度、集约度和空间形态三方面选取城市扩展特征指标,基于自组织网络可视化工具分析上述指标的数值分布和变化规律。自组织网络成分平面图对比分析表明,我国主要大城市在1990-2010年间经历了快速低密度扩张的过程。2010年城市建设用地面积平均扩张为1990年的3. 81倍,而城市建设用地人II密度则平均每年下降3%-4%。与1990-2000年相比较,2000-2010年城市扩张速度更快,人口密度下降更多,低密度扩张特征更为明显,但是扩张最快的城市由一线城市变为二、三线城市。     

Aftershocks hazard in Italy Part I: Estimation of time-magnitude distribution model parameters and computation of probabilities of occurrence

We analyzed the available instrumental data on Italian earthquakes from1960 to 1996 to compute the parameters of the time-magnitudedistribution model proposed by Reasenberg and Jones (1989) andcurrently used to make aftershock forecasting in California. From 1981 to1996 we used the recently released Catalogo Strumentale deiTerremoti `Italiani' (CSTI) (Instrumental Catalog Working Group, 2001)joining the data of the Istituto Nazionale di Geofisica e Vulcanologia(INGV) and of the Italian major local seismic network, with magnituderevalued according to Gasperini (2001). From 1960 to 1980 we usedinstead the Progetto Finalizzato Geodinamica (PFG) catalog(Postpischl, 1985) with magnitude corrected to be homogeneous with thefollowing period. About 40 sequences are detected using two differentalgorithms and the results of the modeling for the corresponding ones arecompared. The average values of distribution parameters (p= 0.93±0.21, Log₁₀(c) = –1.53±0.54, b = 0.96±0.18 and a = –1.66±0.72) are in fair agreementwith similar computations performed in other regions of the World. We alsoanalyzed the spatial variation of model parameters that can be used topredict the sequence behavior in the first days of future Italian seismic crisis,before a reliable modeling of the ongoing sequence is available. Moreoversome nomograms to expeditiously estimate probabilities and rates ofaftershock in Italy are also computed.     

Hydrological modeling using a dynamic neuro-fuzzy system with on-line and local learning algorithm

This paper introduces the dynamic neuro-fuzzy local modeling system (DNFLMS) that is based on a dynamic Takagi–Sugeno (TS) type fuzzy inference system with on-line and local learning algorithm for complex dynamic hydrological modeling tasks. Our DNFLMS is aimed to implement a fast training speed with the capability of on-line simulation where model adaptation occurs at the arrival of each new item of hydrological data. The DNFLMS applies an on-line, one-pass, training procedure to create and update fuzzy local models dynamically. The extended Kalman filtering algorithm is then implemented to optimize the parameters of the consequence part of each fuzzy model during the training phase. Local generalization in the DNFLMS is employed to optimize the parameters of each fuzzy model separately, region-by-region, using subsets of training data rather than all training data.     

太湖梅梁湾沉积岩芯元素地球化学记录的多元统计分析

通过沉积岩芯元素含量垂直分布和富集因子分析,结合多元统计方法如系统聚类和模糊聚类分析,研究了太湖梅梁湾沉积岩芯元素地球化学记录的湖泊环境演化过程.以系统聚类分析和模糊聚类分析为基础,将梅梁湾地球化学记录的环境过程划分为如下几个层段:0-6cm为人类活动强烈干扰的阶段,25-50cm和55-72cm层段都是自然过程的表现,反映了流域物源的影响,而两层段元素记录的差异反映了不同自然作用驱动下的湖泊环境变化;6-25cm和50-55cm层段则是不同环境特征的过渡阶段,因此,综合应用多元统计方法可以更直观更精确地量化影响湖泊环境的因素,有助于恢复湖泊环境历史演化过程.