基于SVM的高耸圆形建筑物沉降数据分析及预报研究

随着高耸圆形建筑物大量修建,高耸圆形建筑物监测成为防治和减少灾害发生的一项极为重要的工作之一。针对小波降噪的平稳特性、小样本条件下支持向量机机器学习和预测的准确性,建立了高耸圆形建筑物的小波去噪、支持向量机机器学习和预测分析模型。实验表明,不确定性支持向量机的预测结果和学习效果的均方差比遗传算法更小,可即时发现高耸圆形建筑物不均匀下沉或倾斜现象。这种数据处理方法经试验表明可以应用到其他预测方面。     

建筑荷载引起漫滩地面沉降数值模拟分析

基于土力学理论,结合数值分析的方法建立长江漫滩典型地质条件下地基土层及其表面的建筑荷载模型,采用ANSYS计算软件对其进行模拟分析,探讨建筑物间距、建筑物规模对建筑物差异沉降的影响,分析相邻建筑物作用下的沉降影响范围变化规律。结果表明,建筑物间距、建筑物规模对建筑物差异沉降会产生不同的影响,相邻建筑物作用下的水平和深度影响范围也有不同的变化规律。     

一种传统民居遥感提取方法

针对基于地物光谱统计特征的建筑物提取方法由于存在较大的同物异谱现象导致提取结果不满足要求的问题,该文提出了一种基于形态学建筑物指数并顾及纹理特征的遥感提取方法。该方法综合考虑传统民居在高分辨率遥感影像上的光谱、形态和纹理特征,首先利用形态学建筑物指数法提取建筑,并使用最小矩形长宽比和像元个数区分道路和零星地物,而后利用Contourlet变换和谱直方图相似性计算进行纹理甄别,实现传统民居的遥感识别和提取。为了验证该方法,选取湖南省常宁市庙前镇中田村QuickBird影像进行试验,结果表明该方法能够获得较高精度的提取结果,整体精度为71.54%,影响提取精度的关键原因为损毁严重的建筑物光谱特征与目标图像纹理相差较大。     

结合机载和地面LiDAR数据的建筑物重建技术

机载LiDAR数据虽然能够快速地获取建筑的顶面信息,但是不能够有效地获取建筑物的侧面信息。地面三维激光扫描仪能够有效地获取建筑物的立面信息,但获取建筑的顶面信息较困难。针对机载和地面LiDAR数据在精细建模中存在的问题,采用机载地面LiDAR数据相结合的方式对建筑物进行精细的建模,实验结果表明,采用该法能够实现建筑物的精细建模。     

基于阈值分割的LIDAR建筑物提取

在介绍常用的阈值分割方法的基础上,详述了基于阈值分割的LIDAR建筑物提取方法,对不同阈值分割方法的优越性与实用性进行分析。研究结果表明,阈值法是一种最简单最基本的图像分割方法。全局阈值能快速有效地分割噪声小、比较均匀的图像,动态阈值和c均值模糊阈值分割对不均匀图像能进行较好的分割,其中c均值模糊阈值分割法最佳。     

利用影像辅助LiDAR点云数据的建筑物提取方法研究

机载LiDAR作为一种新兴的对地观测技术,能够快速地获取地表三维信息。如何从海量LiDAR点云数据中提取建筑物是数据处理中的一项关键工作。本文结合LiDAR数据和航空影像的数据特点,提出了一种航空影像辅助的LiDAR点云建筑物提取方法,首先,采用面向对象方法从航空影像中提取建筑物的轮廓;然后,以建筑轮廓信息为参考,从LiDAR点云中提取建筑物的点云数据;最后,通过实验证明该方法的有效性与可行性。     

面向对象的多源数据建筑物提取

建筑物是城市的重要标志之一,综合利用LiDAR数据和高分辨率遥感影像可以充分发挥不同数据源中提取建筑物的优势。本文基于面向对象分类理论,利用机载LiDAR数据和GeoEye高空间分辨率遥感影像,在多尺度分割的基础上对实验区分类并提取建筑物,进而对提取结果进行精度评价。实验表明,将LiDAR数据与高分辨率影像数据结合能够很好地提取建筑物,建筑物提取精度达89.28%。     

Folk housing revisited

As Fred Kniffen observed, vernacular buildings identify culture and record our relationships with physical and social environment. Influenced by Kniffen, twentieth‐century cultural geographers used spatially correlated log homebuilding attributes as diagnostics. The present study used a qualitative meta‐study approach to evaluate studies citing such correlations in the eastern temperate forests of North America. Forty‐two studies involving sixty‐three geographic entities and twenty‐two attribute types were evaluated. The meta‐study's findings were consistent with an Eastern Woodlands regional model described by Kniffen, Terry Jordan, and Wilbur Zelinsky. A majority of the spatially correlated attributes involved building materials, cited cultural and/or environmental influences to explain their findings, and cited correlations at state/province or county scales. Today, identification of building culture undoubtedly continues to offer potential guidance to sustainability efforts, and, although untapped, vernacular building continues to offer potential as a key diagnostic.     

Policy making for renewable energy in India: lessons from wind and solar power sectors

It is clear that developing countries will have to be part of the global mitigation effort to avoid ‘dangerous climate change’, and, indeed, many of them are already undertaking significant actions on multiple fronts to help address this problem, even if they have not yet taken on legally binding commitment under the United Nations Framework Convention on Climate Change (UNFCCC). Since the deployment of GHG-mitigating technologies is already a significant part of this effort and likely to be even more so in the future, drawing lessons from existing programmes can help accelerate and enhance the effectiveness of this deployment process. Accordingly, this article aims to examine the deployment of wind and solar power in India, paying specific attention to the role of public policy in incentivizing and facilitating this deployment, how these policies have evolved over time, what has shaped this evolution, and what the learning has been over this period. Through this analysis, the intention is to draw out key lessons from India's experience with deployment policies and programmes in these two sectors and highlight the issues that will need to be given particular consideration in the design of future domestic policies and international cooperation programmes to enhance the move towards climate-compatible development in India. Many of these lessons should also be relevant for other developing countries that are attempting to balance their climate and developmental priorities through the deployment of renewable energy technologies.     

结合影像的LIDAR数据三维建筑物提取

随着LIDAR技术的出现,三维建筑物的提取也受到越来越多的重视。由于LIDAR数据分布的不连续性和不规律性,直接从机载激光扫描测距数据中进行建筑物提取较为困难。本文提出了一种结合灰度影像的LIDAR数据三维建筑物提取方法,分三个步骤:首先对灰度影像进行建筑物二维提取;然后将处理后影像和LIDAR数据粗匹配,初步确定LI-DAR数据中的三维建筑物区域;最后利用一组阈值操作进行三维建筑物的精确提取。实验结果表明该方法简单实用,适应性强。