基于分类与形态综合的高分辨率影像建筑物提取方法研究

本文提出了一种基于分类和形态综合的高分辨率影像建筑物提取方法。主要研究的是基于高分辨率影像多尺度分割的面向对象人工地物分类方法,并以此为基础提取房屋,实现房屋位置、范围的初步自动化提取。最后,对提取的地物图斑的精确边界利用Hough变换提取直线边缘轮廓线,并以此为基础进行轮廓拟合和制图综合,从而得到规则的、方形组合的建筑物形状信息。     

从卫星遥感全色图像中自动提取城市目标

提出了一种从卫星遥感全色图像中自动提取城市目标的方法。在总结城市目标的卫星遥感全色图像特性的基础上,设计了一种层次提取的方法。首先在低分辨率图像中提取城市目标的候选区域;接着在候选区域内利用城市目标一些更复杂的特性在高分辨率图像中证实城市目标。为了获得城市目标的边界,提出了一种基于边缘点密度的算法,并将算法应用去提取SPOT图像中的城市目标,取得了较好的试验结果。     

白沙河地下水资源管理模型

针对白沙河地下水含水层的内部结构和本区开采特点以及已出现的环境问题，建立了地下水资源管理模型，确定了优化开采方案。该管理模型针对本区每年２个管理时段的长度及其开采量均不同的特点，导出了新的响应矩阵的公式。管理模型的建立有助于对该区地下水的合理开采。     

集成高度和彩色纹理特征的影像目标模糊聚类识别方法

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。     

视频对象分割及跟踪方法研究

讨论和分析了视频对象全自动提取及后续帧中的跟踪过程及采用的方法,对其算法、分割性能和结果进行了比较和评述。并结合Hausdorff和Snake跟踪器,对初始对象轮廓进行跟踪。结果证明,这种跟踪方法能有效地解决对象被遮挡和形变以及背景移动的问题。     

线与线之间的空间拓扑关系组合推理

在空间拓扑关系组合描述的基础上,进一步完善了利用基本空间拓扑关系进行组合推理的方法,并建立了组合表,详细绘制了两条线之间的空间拓扑关系图。     

一种计算图斑净面积的新方法

在图斑净面积计算规则下,对地类图斑和线状地物两个图层的组织结构和相互关系展开研究,提出Arc-Map空间分析计算图斑净面积方法。采用实例验证了该方法的可行性和准确性,通过对比分析可以看出,Arc-Map空间分析法具有良好的使用和推广价值。     

A geometric-based approach for road matching on multi-scale datasets using a genetic algorithm

Object matching is used in various applications including conflation, data quality assessment, updating, and multi-scale analysis. The objective of matching is to identify objects referring to the same entity. This article aims to present an optimization-based linear object-matching approach in multi-scale, multi-source datasets. By taking into account geometric criteria, the proposed approach uses real coded genetic algorithm (RCGA) and sensitivity analysis to identify corresponding objects. Moreover, in this approach, any initial dependency on empirical parameters such as buffer distance, threshold of spatial similarity degree, and weights of criteria is eliminated and, instead, the optimal values for these parameters are calculated for each dataset. Volunteered geographical information (VGI) and authoritative data with different scales and sources were used to assess the efficiency of the proposed approach. According to the results, in addition to an efficient performance in various datasets, the proposed approach was able to appropriately identify the corresponding objects in these datasets by achieving higher F-Score.     

Mostly Dormant Comets and their Disintegration into Meteoroid Streams: A Review

The history of associating meteor showers with asteroidal-looking objects is long, dating to before the 1983 discovery that 3200 Phaethon moves among the Geminids. Only since the more recent recognition that 2003 EH1 moves among the Quadrantids are we certain that dormant comets are associated with meteoroid streams. Since that time, many orphan streams have found parent bodies among the newly discovered Near Earth Objects. The seven established associations pertain mostly to showers in eccentric or highly inclined orbits. At least 35 other objects are tentatively linked to streams in less inclined orbits that are more difficult to distinguish from those of asteroids. There is mounting evidence that the streams originated from discrete breakup events, rather than long episodes of gradual water vapor outgassing. If all these associations can be confirmed, they represent a significant fraction of all dormant comets that are in near-Earth orbits, suggesting that dormant comets break at least as frequently as the lifetime of the streams. I find that most pertain to NEOs that have not yet fully decoupled from Jupiter. The picture that is emerging is one of rapid disintegration of comets after being captured by Jupiter, and consequently, that objects such as 3200 Phaethon most likely originated from among the most primitive asteroids in the main belt, instead. They too decay mostly by disintegration into comet fragments and meteoroid streams. The disintegration of dormant comets is likely the main source of our meteor showers and the main supply of dust to the zodiacal cloud. Editorial handling: Frans Rietmeijer.     

面向对象的连云港海岸带土地利用变化及驱动力分析

以连云港海岸带为研究区,基于1978年和2010年两期Landsat影像,提取土地利用信息,分析土地利用特征及其动态变化过程;面向土地利用变化基本单元,进行土地利用变化驱动力分析。结果表明:研究期初,该地区的主要土地类型为耕地和盐田,共占研究区面积的67.14%;32 a间共有37.17%的土地发生了变化,盐田和耕地面积快速减少,水产养殖区和工矿企业用地快速增加,有限的林地资源在减少,滩涂围垦达面积46.55 km2,土地利用程度加深;变化的土地中,38.95%由经济结构调整驱动,发展潜力大;27.75%由政策驱动,存在诸多的社会和生态问题;20.49%由政策或者经济结构调整与生产力水平共同驱动,取得较好经济和社会效益;12.81%的土地由政策、人口及经济发展共同驱动。连云港海岸带土地利用存在诸多不足之处,也具有巨大的发展空间。本研究对连云港市提高土地利用的社会效益和生态效益具有重要意义。