Bayesian and Geostatistical Approaches to Combining Categorical Data Derived from Visual and Digital Processing of Remotely Sensed Images

This paper seeks a synthesis of Bayesian and geostatistical approaches to combining categorical data in the context of remote sensing classification. By experiment with aerial photographs and Landsat TM data, accuracy of spectral, spatial, and combined classification results was evaluated. It was confirmed that the incorporation of spatial information in spectral classification increases accuracy significantly. Secondly, through test with a 5-class and a 3-class classification schemes, it was revealed that setting a proper semantic framework for classification is fundamental to any endeavors of categorical mapping and the most important factor affecting accuracy. Lastly, this paper promotes non-parametric methods for both definition of class membership profiling based on band-specific histograms of image intensities and derivation of spatial probability via indicator kriging, a non-parametric geostatistical technique.

Fracture network characterisation of the Balmuccia peridotite using drone-based photogrammetry,implications for active-seismic site survey for scientific drilling

The presence of discontinuities(e.g.faults,fractures,veins,layering)in crystalline rocks can be challenging for seismic interpretations because the wide range of their size,orientation,and intensity,which controls the mechanical properties of the rock and elastic wave propagation,resulting in equally varying seismic responses at different scales.The geometrical characterisation of adjacent outcrop discontinuity networks allows a better understanding of the nature of the subsurface rocks and aids seismic interpretation.In this study,we characterise the discontinuity network of the Balmuccia peridotite(BP)in the IvreaeVerbano Zone(IVZ),northwestern Italy.This geological body is the focus of the Drilling the Ivrea eVerbano zonE(DIVE),an international continental scientific drilling project,and two active seismic surveys,SEismic imaging of the Ivrea ZonE(SEIZE)and high-resolution SEIZE(Hi-SEIZE),which aim to resolve the subsurface structure of the DIVE drilling target through high-resolution seismic imaging.For fracture characterisation,we developed two drone-based digital outcrop models(DOMs)at two different resolutions(10^(-3)-10 m and 10^(-1)-10^(3)m),which allowed us to quantitatively characterise the orientation,size,and intensity of the main rock discontinuities.These properties affect the seismic velocity and consequently the interpretation of the seismic data.We found that(i)the outcropping BP discontinuity network is represented by three more sets of fractures with respect to those reported in the literature;(ii)the discontinuity sizes follow a power-law distribution,indicating similarity across scales,and(iii)discontinuity intensity is not uniformly distributed along the outcrop.Our results help to explain the seismic behaviour of the BP detected by the SEIZE survey,suggesting that the low P-wave velocities observed can be related to the discontinuity network,and provide the basic topological parameters(orientation,density,distribution,and aperture)of the fracture network unique to the BP.These,in turn,can be used for interpretation of the Hi-SEIZE seismic survey and forward modelling of the seismic response.

MAPPING OF OPHIOLITES IN THE INDUS-SUTURE ZONE OF NORTHWESTERN HIMALAYA, USING SATELLITE REMOTE SENSING DATA

Ophiolites, which have been tectonically emplaced along continental margins and island arcs, are significant to the understanding of mountain belt evolution. In the Himalayas, the ophiolitic suite of rocks occur along the Indussuture zone from Hanle in the southeast to Dras\|Kargil sector in the northwest and it represents the remnant of the compressed uplifted wedge of the oceanic crust between the two colliding continental masses, the Indian and the Asian plates.. These ophiolites are temporally and spatially correlated with the culminating phase of the Himalayan orogeny. The Indus River flows to its north separating the ophiolite from the Trans Himalayan litho\|units. Geological mapping in the hostile and inaccessible mountainous terrains of the Himalaya has always posed a great challenge to geologists. Nevertheless, a number of geologists have undertaken such arduous mapping expeditions in the past and prepared fairly good geological maps of these terrains .However there always existed disputes on the accuracy of lithological boundaries and structural details in these maps because many of these boundaries and structural features were completed through extrapolations and/or interpolations as the ruggedness and inaccessibility of a large part of the terrain forbid physical examination of every outcrop. It is in this context the potential of remote sensing, especially of satellite images, is to be appreciated.

THE APPLICATION OF REMOTE SENSING TECHNIQUE ON GEOLOGICAL INVESTIGATION OF PLACER DEPOSIT

The practice has proved that it is an economic and effective method to investigate placer gold deposit by using multi-level information sources of remote sensing and multi-variate analysis methods, especially for the area with a sparse population and difficult condition like the Da Hinggan Mountains, China.The information sources used in our work includes Landsat TM, aerial infrared photography and their mosaic image maps and enlarged photos with different scales. According to statistic data, in the study area the gold-bearing rocks are mainly granite, alaskite, granodiorite and some old metamorphic rocks. On gold-bearing geological structures, the fault zones in the four directions (NE, NNE, NW and EW) are obvious, in which NNE and EW are the most key fault zones. On fluvial geomorphology the flow courses stored placer are in the tributaries of the 4th and 5th levels, especially in straight or slight curve reaches. On the basis of analysis the interpretative signs were set up, and the interpretative

Research on a Watermarking Algorithm for Remote Sensing Images Resist Screen Capture Attacks in Urban Planning Information Management

Remote Sensing data, as an essential urban basic information in urban planning, has the characteristics of large information capacity, real time, high update speed and accuracy. Because of urban spatial information involving multi-faceted public and public interests, its data security is very important. The use of digital watermarking technology can effectively protect the secu-rity of urban planning basic data. In practical applications, the “screen capture” poses a great threat to the security of remote sensing image. In order to resist the screen capture attacks, the QR code watermark information is encoded and converted, and combined with a circular angle template watermark, a digital watermarking algorithm for remote sensing images in urban planning information management is proposed. And the proposed algorithm is experimentally verified. Experiments show that the algorithm is robust against screen capture attacks, and provide security guarantee for urban construction and management.

Photogrammetry - Remote Sensing on the Study of Monuments and Historical Centers： The Effect of Hazards -The Case of Delphi Historical Center

Monuments and historical centers, because of their particular importance, are studied in multiple ways. The study concerns different scientific disciplines and technology. Photogrammetry and remote sensing contribute essentially to this study, because of the valuable qualitative and quantitative information they offer. In this paper we search through the possibilities of very high resolution satellite imagery on historical centers study, referring to Delphi historical center. The study concerns image enhancement techniques and visual interpretation of Ikonos satellite imagery. Image enhancement techniques facilitate visual interpretation, detection and recognition, of the physiognomy and spatial arrangement of Delphi historical center and offer information about physical and architectural features in the wide area of the historical center.

Applying Digital Image Processing to Evaluate a Extraction Method of Cartographic Features in Digital Images

A topic studied in cartography is to make the extraction of cartographic features that provide the update of cartographic maps more easily. For this reason many automatic routines were created with the intent to perform the features extraction. Despite of all studies about this, some features cannot be found by the algorithm or it can extract some pixels unduly. So the current article aims to show the results with the software development that uses the original and reference image to calculate some statistics about the extraction process. Furthermore, the calculated statistics can be used to evaluate the extraction process.

The Digital Mapping of Ukrainian Soils on the Base of High Resolution Space Images

The first Ukrainian using experience of multispectral space scanning for digital soil mapping is described in this paper. Methodical approaches for detailed soil observation of Ukrainian forest regions are elaborated based on modem mapping principles. For the first time in Ukraine, digital soil maps based on GIS （geographic information system） were obtained for individual farms. In GIS based on space images and digital relief models, the medium-scale and large-scale soil maps were created by geo-statistical methods. According to elaborated methods, modem digital soil mapping should provide all combined works： remote sensing and traditional soil observations. The modem digital soil mapping should be based just on quantitative principles： on remote sensing data, geomorphologic field parameters, and chemical analyses. The methodological approaches, which were used for the first time in Ukraine during digital soil mapping by remote sensing methods, are described in this paper.

基于改进HRNet的遥感影像冬小麦语义分割方法

冬小麦在影像中呈现田块碎小且分布零散等空间特征,同时影像包含的复杂地物对冬小麦识别造成干扰,易出现识别精度低且边界分割模糊等问题。为及时准确获取大范围冬小麦空间分布信息,该研究以高分二号卫星影像作为数据源,提出一种CAHRNet(change attention high-resolution Net)语义分割模型。采用HRNet(high-resolution Net)替换ResNet作为模型的主干网络,网络的并行交互方式易获取高分辨率的特征信息;联合OCR(object-contextual representations)模块聚合上下文信息,以增强像素点与目标对象区域的关联性;3)引入坐标注意力(coordinate attention)机制,使网络模型充分利用有效的空间位置信息,以保留分割区域的边缘细节,提高对分布零散、形状多变的冬小麦田块的特征提取能力。试验结果表明,在自制的高分辨率遥感数据集上,CAHRNet模型的平均交并比(mean intersection over union,MIoU)和像素准确率(pixel accuracy, PA)分别达到81.72%和97.08%,MIoU相较U-Net、DeepLabv3+分别提高了9.09、2.44个百分点;PA相较U-Net、DeepLabv3+分别提高6.80、1.59个百分点,说明CAHRNet模型具有较高的分割识别精度,可为进一步准确获取冬小麦作物分布信息提供技术支撑。

拉普拉斯卷积的双路径特征融合遥感图像智能解译方法

由于遥感图像存在多尺度变化和目标边缘模糊等问题,对其进行智能解译仍然是一项极具挑战性的工作。传统的语义分割方法在处理这些问题时存在局限性,难以有效捕捉全局和局部信息。针对上述问题,文中提出一种双路径特征融合分割方法 DFNet。首先,使用Swin Transformer作为主干提取全局语义特征,以处理像素之间的长距离依赖关系,从而促进对图像中不同区域相关性的理解;其次,将拉普拉斯卷积嵌入到空间分支,以捕获局部细节信息,加强目标地物边缘信息表达;最后,引入多尺度双向特征融合模块,充分利用图像中的全局和局部信息,以增强多尺度信息的获取能力。在实验中,使用了三个公开的高分辨率遥感图像数据集进行验证,并通过消融实验验证了所提模型不同模块的作用。实验结果表明,所提方法在Uavid数据集、Potsdam数据集、LoveDA数据集的mIoU达到了71.32%、85.58%、54.01%,提高了语义分割的性能,使分割结果更为精细。

不同影像处理方法在大坝施工范围识别中的应用分析

【目的】施工范围是大坝建设过程中施工精细化管理的基础内容,传统的施工范围识别方法主要是人工解译,效率低、成本高、主观性强。将影像智能解译的新方法应用于大坝施工范围识别成为一项有挑战性的技术问题。【方法】以我国贵州省凤山水库为研究区,通过对大坝施工范围特点与现有常用的遥感影像解译方法的分析,针对大坝施工范围具有的区域面积大且连续、地物复杂、人工扰动性程度高等特征,选用高分二号和哨兵二号遥感影像,研究不同的遥感影像解译方法在大坝施工范围自动识别的应用效果。【结果】文中用到的三种方法均取得较好的结果,且分别具有不同的优势。【结论】结果表明,基于影像分割的施工范围识别方法、基于变化检测的施工范围识别方法和基于影像智能分类的施工范围识别方法均能取得较好的识别结果,其中第二种方法最优,且无需人工干预;第三种方法次之,但是方法二和方法三对样本的数量和质量依赖性较高,且存在离散图斑错分的现象;第一种方法容易受阈值的选择和叠加分析的影响,识别精度低于后两种,但是处理流程简单、容易实现。

“任务链+研学支架”教学模式的建构——新工科视角下“遥感数字图像处理”课教学改革

目前,传统教学在内容的系统性、认知活动高阶性、价值引领实效性、促学效果等方面存在着较严重的问题,为此,“遥感数字图像处理”课程基于新工科教学理念,以遥感技术的综合应用情境重构课程体系,开启“任务链+研学支架”教学模式,加强培养学生遥感应用能力和创新能力,以遥感技术服务社会需求为育人路径。新模式的构建,实现了应用与研学互动、研学与研教并举、育人与育才同构的效果,对同类型新工科课程具有借鉴意义。

遥感专业《数字图像处理》教材建设有效途径分析

数字图像处理教材存在理多实少、难点解析不足等种种问题。针对这些问题,提出一种混合式立体改进方法,从内容、形式和实践方面提出了多项举措,以实现够用学会、讲练透彻的学习目标。通过分析遥感专业课程内容,提炼出相关知识点进行教材编写,得到理实一体的简明化教材。教学实践证明,这种教材建设方法丰富了教学资源,有利于学生对重难点知识理解,大幅提高学习效率和质量,不失为一种有效的逆向教材建设创新方法。

基于卷积神经网络的光学遥感影像道路提取方法研究进展

随着光学遥感影像空间分辨率的提升和获取渠道的丰富,利用光学遥感影像实现地物智能解译已成为高效的技术路径。由于卷积神经网络(convolutional neural networks,CNN)强大的特征提取能力以及道路信息在多个领域的应用需求,基于CNN的道路提取方法成为了当前的研究热点。鉴于此,本文根据近年来的相关研究文献,对基于CNN的道路提取方法从基于形状特征的改进、基于连通性的改进、基于多尺度特征的改进和基于提取策略的改进四个方面进行归纳总结,然后描述典型道路遮挡案例,并利用经典CNN从样本标签的局限性层面对当前的技术难点进行分析与验证,最后从多源数据协同、样本库建设、弱监督模型和域适应学习四个方面对遥感影像道路提取的发展趋势进行评估和展望。

无序无人机影像的并行化SfM三维重建方法

增量式运动恢复结构(ISfM)已成为无人机影像三维重建的关键技术。然而,大数据量、大重叠度和高分辨率的无序无人机影像导致的匹配对检索代价大、迭代优化误差累积和效率低的问题,使其难以满足大场景ISfM处理需求。本文提出联合全局描述子和图索引的无人机影像并行化SfM方法。针对影像特征数量大、影像检索编码本尺寸增加导致的匹配对检索效率低的问题,设计了联合全局描述子和图索引的高效影像检索方法,从而加速影像匹配。针对分块并行化SfM子场景合并存在同名点搜索效率低、内存消耗大、合并解算精度低的问题,设计了基于按需匹配图和双向重投影误差的子场景合并方法,实现无人机影像的并行化SfM重建。利用不同场景、不同采集方式获取的真实无人机影像进行试验,结果表明本文方法能够实现36~108倍加速比的匹配对检索,ISfM重建效率达到30倍加速,且相对定向和绝对定向精度与传统方法相当。

遥感图像超分辨率重建技术研究及地质解译应用

遥感图像超分辨率重建技术通过提升遥感图像的空间分辨率,增强图像的细节和清晰度,在地质调查中能够提高地质特征的识别精度和信息提取的准确性,具有重要的应用价值。针对现有遥感图像重建模型在地质解译下的细节和特征信息丢失等缺点,提出一种基于增强型深度残差网络的多尺度渐进式残差网络(MPEDSR)的超分辨率重建模型。该模型重点针对残差结构进行调整,在残差块堆叠结构中加入跳跃连接,同时在残差块间引入通道注意力机制模块,实现残差分支信息的充分利用,并通过测试集图像证明所提方法的地质调查要素解译能力。实验结果表明:所提算法在主观视觉感受及峰值信噪比(PeakSignal-to-NoiseRatio,PSNR)和结构相似度(Structural Similarity,SSIM)等客观评价指标上均高于现有几种主流算法。在2倍和4倍放大因子下相较原模型在PSNR上分别提升0.176和0.194 dB,在SSIM上分别提升0.018和0.021。此外,该网络能有效提取滑坡和岩性边界,识别断裂构造。本研究为高分辨率遥感图像下的地质解译和地质灾害监测提供了有效的技术手段,推动了遥感地质解译的精细化和智能化发展。

高分辨率遥感影像样本库动态构建与智能解译应用

在人工智能时代,遥感影像解译朝着自动化智能化方向发展,高质量的样本数据集是其核心。我国积累了海量优质的时空地理信息基础数据及衍生产品,是深度学习驱动的遥感影像智能解译样本的重要来源。盘活现有数据资源,可推动人工智能与遥感解译的应用深度与广度。本文基于现有数据资源,针对样本数据集区域受限、时效性不强、类型单一等问题,研究了面向深度学习的高分遥感影像智能解译样本库动态构建技术。首先,分析了要素提取、地表覆盖分类、变化检测方面的公开样本数据集的特点,提出业务驱动的样本应需生成-动态构建-智能应用思路;其次,研究了基于历史解译成果的样本自动生成、SAM大模型提示学习引导的样本清洗精化方法及实现过程;再次,设计了具有区域性、时序性、尺度性、多传感器、多类型的样本库,以及顾及空间-时间-地类关系的动态样本数据库架构,研究了样本数据集“量化-检索-组合”动态重构过程,实现时空样本的动态管理与多维检索;最后,开展了地表覆盖分类、要素提取、变化检测等智能解译应用,验证了本文研究思路及方法的可行性,以期推动基于已有基础数据的样本数据集的有效利用,以及样本构建-管理-应用及数据-模型-业务的互联互通,为高分遥感影像智能解译样本库构建与应用提供参考思路。

基于高分辨率遥感影像的土耳其地震建筑物震害特征

选取土耳其卡赫拉曼马拉什省努尔达吉镇城区和周边乡村地区作为研究区,通过获取地震前后的高空间分辨率遥感影像数据,提取房屋建筑震害信息,识别房屋建筑震害等级,利用震害指数计算区域内地震烈度等级,综合分析得到建筑物震害特征与主要成因,对地震灾情快速评估、应急救援和恢复重建具有重要参考。

基于无人机影像的沙漠植被生物量反演研究

利用无人机遥感影像,采用随机森林算法结合面向对象方法对沙漠植被进行遥感解译并分类,利用特定植被株高、冠幅反演模型对沙漠植被生物量进行定量分析,对沙漠植被生物量研究、沙漠植被分布状况研究、沙漠植被生长状态研究等具有一定的借鉴作用。