Comparative assessment of semantic-sensitive satellite image retrieval: simple and context-sensitive Bayesian networks

In recent years, Bayesian networks using unsupervised extracted image features have been applied in many remote sensing information mining systems to enable semantic-sensitive image retrieval. However, a simple Bayesian network insufficiently accounts for the spatial information, that is, the relations among image regions, for the semantic inference process. This drawback significantly impacts the retrieval performance, especially if the utilised features contain no or little spatial information. Therefore, this article proposes a context-sensitive Bayesian network, which infers semantic concepts of image regions based on the spectral and textural characteristics of the regions themselves as well as their contexts, that is, the adjacent regions. In order to compare the context-sensitive Bayesian network with the simple Bayesian network, comprehensive experiments were conducted based on high-resolution multispectral IKONOS imagery. The results show that the incorporation of the image regions' spatial relations not only significantly improves the accuracy of the semantic concepts inference, but also allows more flexibility in choosing the type of low-level features.     

Extraction of Lithium from Salt Lake Brine using N523-TBP Mixture System

In this work,problems encountered by tri-butyl phosphate( TBP) in the industrialization of lithium extraction from salt lake brine were discussed in detail. The behavior of N,N-bi-( 2-ethylhexyl) acetamide( N523) during lithium extraction was investigated,and its disadvantages were analyzed in the view of practical application. An N523-TBP mixture extraction system was proposed to alleviate or avoid the defects that N523 and TBP met when they were used separately. The optimal composition of this mixture extraction system was determined to be 20% N523-30% TBP-50% kerosene.The effects of brine acidity,Fe/Li molarity ratio,phase ratio and chloride ion concentration on lithium extraction efficiency were discussed. The operation conditions in single-stage extraction were optimized as brine acidity = 0. 05 mol/L,Fe/Li molarity ratio = 1. 3,and phase ratio = 2. The high concentration of chloride ions in brine was beneficial for the extraction of lithium. The structure of the extracted complex was proposed as( LiFeCl_4·n N523·m TBP)·( 2-n) N523·( 2-m) TBP( m + n = 2) by chemical analysis and slope-fitting methods. The extraction thermodynamic functions were calculated preliminarily,and the results suggested that the extraction process was an exothermic( ΔH 0) and spontaneous( ΔG 0) reaction,and the degree of disorder increased( ΔS 0) during the extraction process. This work will give some guidance to the lithium industry of Qinghai in both the fundamental theory and practical application.     

贵州织金片区影像地层研究

选定遥感解译程度较好的贵州西部高原山区织金片作为研究对象,以美国Landsat/TM数据为数据源,选择各波段之问相关系数最小的TM7、TM4、TM1波段合成假彩色图像,采用人机交互的目视解译方法进行影像地层划分.研究区碳酸盐岩地层以其独特的溶蚀地貌而显示为斑状、块状影像,地表水系不发育;碎屑岩地层因流水侵蚀地貌而表现为曲线状、网状纹理,从而使不同地层上下层位在空间结构信息上显示出差异.据此,通过界面、特殊纹理、代表性色调和典型水系等对地层影像进行比较、确认.在此基础上,将研究区主要出露的地层划分为14个影像地层单元.经野外验证,影像地层划分结果与传统的地层划分单元或单元组合吻合度较好.     

高分一号遥感影像地质灾害信息提取方法研究

山区地质灾害发生频繁,受灾范围大,危害严重。提取地质灾害信息、估算受灾情况,对救灾工作极为重要。传统的遥感人工解译方法速度慢、效率低;计算机自动解译速度快但解译精度受影像质量影响大,大范围地区很难建立起普适性的解译模型。本文利用洮南市西部高分一号遥感数据,结合DEM生成三维影像,建立解译标志,解译基础地质信息,分析地质灾害成因;根据研究区泥石流地质灾害的主要影像特征,提取影像的分类属性,基于面向对象的分类方法,建立信息提取模型,快速提取出地质灾害敏感区域;再进行人机交互,精确提取出地质灾害的类型和范围,通过野外验证,该方法十分可靠,为大范围地质灾害信息快速提取和灾后救援提供科学依据。     

共线条件方程在建筑物比高求解中的应用

高度信息作为人工地物的重要属性信息,在军事和民用上都具有很高的利用价值。针对人工地物高度提取的发展现状及存在的主要问题,对单张航片利用共线条件方程确定人工地物比高方法的原理和技术进行了研究,并进行了实验,验证了方法的可行性。     

空-谱融合的条件随机场高光谱影像分类方法

高光谱遥感数据具有光谱信息丰富、图谱合一的特点,目前已经广泛地应用在对地观测中。传统的高光谱分类模型大多过分依赖影像光谱信息,没有充分利用空间特征信息,这使得分类精度还有很大的提升空间。条件随机场是一种概率模型,能够较好地融合空间上下文信息,在高光谱影像分类中已经得到越来越多的关注,但大部分条件随机场模型存在超平滑的现象,会导致影像细节丢失。针对该问题,本文提出了一种优化融合影像空-谱信息的高分辨率/高光谱影像分类方法,该方法将影像的纹理信息与原始光谱信息进行融合,利用SVM分类器对其进行预分类,并将各类概率定义为一元势函数,以融合空间特征信息;然后将空间平滑项和局部类别标签成本项加入二元势函数中,以考虑空间背景信息,并保留各类别中的详细信息。最后,通过两组的高分辨率/高光谱影像数据进行试验。结果表明,与SVM算法、传统的条件随机场方法和面向对象的分类方法相比,本文提出的算法在整体分类精度上分别提高了10%、9%和8%以上,同时在保持地物边缘完整性、避免“同谱异物”与“同物异谱”的现象方面有较明显的优势。     

Smart 3D数据的三维模型重建

针对Smart 3D生成的实景三维模型是一个三角网集合,并不能直接应用于数字城市建设、土地利用管理与城市规划的问题,该文提出了Smart 3D数据三维模型重建的方法:研究了基于分割过滤方法,提取高精度数字高程模型;基于参数化建模、纹理重建方法构建了建筑物三维模型。利用敦化市的倾斜影像进行了实验。实验结果表明:三维重建模型精度高、效果好、提高了自动化的程度,实现了城市建筑物模型的单体化。     

单幅遥感影像三维重建辅助加密DEM数据研究

介绍了一种基于单幅遥感影像三维重建的DEM数据加密方法,提出了用单幅高分辨率遥感影像三维重建方法获得高程格网内地表形态,实现了获取更高分辨率的数字高程模型.     

插拔桩靴对临近桩靴承载力的影响研究

自升式钻井船桩靴的插拔过程将会对临近井口平台的桩靴基础产生不利影响。开展了室内模型试验研究,将试验结果与理论分析结果相结合,得到了桩靴插拔的影响范围及影响范围内土体强度的变化规律,并推导了桩靴插拔影响区内临近桩靴承载力的计算方法,与模型试验结果进行了对比验证。研究表明,桩靴插拔后砂土相对密实度降低,其降低程度随距桩靴距离的增加而逐渐减小,影响范围约为1倍桩靴直径。受桩靴插拔影响,临近桩靴的承载力下降,降低比例随两桩靴相对间距的增大而减小,随桩靴相对插深的增加而增加。当两桩靴相对间距大于2倍插拔桩靴直径时,可认为桩靴插拔对临近桩靴基础承载力无影响。     

一种建立在CPN基础上的分类方法在遥感影像分类中的应用研究

对偶神经网络利用了自组织映射近似函数的一种新的映射神经网络,其结构组合了Kohonen的自组织映射和Grossberg的外星(Outstar)结构,网络结构相对简单。本文以对偶神经网络分类方法原理为基础,研究了一种理想化的分类方法,并以MATLAB平台为基础对遥感影像进行分类处理,实验结果表明,其分类总精度为94.17%,分类精度较传统监督分类结果有所提高。