THE ANALYSIS OF REMOTE SENSING IMAGES FOR ACTIVE FAULTS AND EARTHQUAKES IN CHINA

The Landsat image information has recently been widely applied to structural geology, especially to the analysis of lineaments, owing to their macroscopic, visual and comprehensive features. The images will be more effective when applied to the interpretation of active faults. Active faults are widely ditributed in China. Much attention has been paid to the study of active faults both in China and abroad. There is certain controversy concerning the implication of the term "active fault". Strictly speaking, the term should refer only to the faults that are still active in the present day. However, the term also usually refers to the faults which have been active continually or intermittently from the Quaternary (or the end of Tertiary) to the present day. We propose that the tones and the configurations of features on Landsat images are the principal keys to the interpretation of active faults. The faults, which display the most prominent

Privacy‐preserving remote sensing images recognition based on limited visual cryptography

With the arrival of new data acquisition platforms derived from the Internet of Things(IoT),this paper goes beyond the understanding of traditional remote sensing technologies.Deep fusion of remote sensing and computer vision has hit the industrial world and makes it possible to apply Artificial intelligence to solve problems such as automatic extraction of information and image interpretation.However,due to the complex architecture of IoT and the lack of a unified security protection mechanism,devices in remote sensing are vulnerable to privacy leaks when sharing data.It is necessary to design a security scheme suitable for computation‐limited devices in IoT,since traditional encryption methods are based on computational complexity.Visual Cryptography(VC)is a threshold scheme for images that can be decoded directly by the human visual system when superimposing encrypted images.The stacking‐to‐see feature and simple Boolean decryption operation make VC an ideal solution for privacy‐preserving recognition for large‐scale remote sensing images in IoT.In this study,the secure and efficient transmission of high‐resolution remote sensing images by meaningful VC is achieved.By diffusing the error between the encryption block and the original block to adjacent blocks,the degradation of quality in recovery images is mitigated.By fine‐tuning the pre‐trained model from large‐scale datasets,we improve the recognition performance of small encryption datasets for remote sensing images.The experimental results show that the proposed lightweight privacy‐preserving recognition framework maintains high recognition performance while enhancing security.

Roughness correction method for salinity remote sensing using combined active/passive observations

Roughness-induced emission from ocean surfaces is one of the main issues that affects the retrieval accuracy of sea surface salinity remote sensing.In previous studies,the correction of roughness effect mainly depended on wind speeds retrieved from scatterometers or those provided by other means,which necessitates a high requirement for accuracy and synchronicity of wind-speed measurements.The aim of this study is to develop a novel roughness correction model of ocean emissivity for the salinity retrieval application.The combined active/passive observations of normalized radar cross-sections(NRCSs)and emissivities from ocean surfaces given by the L-band Aquarius/SAC-D mission,and the auxiliary wind directions collocated from the National Centers for Environmental Prediction(NCEP)dataset are used for model development.The model is validated against the observations and the Aquarius standard algorithms of roughness-induced emissivity correction.Comparisons between model computations and measurements indicate that the model has better accuracy in computing wind-induced brightness temperature in the upwind/downwind directions or for the surfaces with smaller NRCSs,which can be better than 0.3 K.However,for crosswind directions and larger NRCSs,the model accuracy is relatively low.A model using HH-polarized NRCSs yields better accuracy than that using VV-polarized ones.For a fair comparison to the Aquarius standard algorithms using wind speeds retrieved from multi-source data,the maximum likelihood estimation is employed to produce results combining our model calculations and those using other sources.Numerical simulations show that combined results basically have higher accuracy than the standard algorithms.

Extraction of coastline in high-resolution remote sensing images based on the active contour model

While executing tasks such as ocean pollution monitoring,maritime rescue,geographic mapping,and automatic navigation utilizing remote sensing images,the coastline feature should be determined.Traditional methods are not satisfactory to extract coastline in high-resolution panchromatic remote sensing image.Active contour model,also called snakes,have proven useful for interactive specification of image contours,so it is used as an effective coastlines extraction technique.Firstly,coastlines are detected by water segmentation and boundary tracking,which are considered initial contours to be optimized through active contour model.As better energy functions are developed,the power assist of snakes becomes effective.New internal energy has been done to reduce problems caused by convergence to local minima,and new external energy can greatly enlarge the capture region around features of interest.After normalization processing,energies are iterated using greedy algorithm to accelerate convergence rate.The experimental results encompassed examples in images and demonstrated the capabilities and efficiencies of the improvement.

An improved algorithm for retrieving thin sea ice thickness in the Arctic Ocean from SMOS and SMAP L-band radiometer data

The aim of this study was to develop an improved thin sea ice thickness(SIT)retrieval algorithm in the Arctic Ocean from the Soil Moisture Ocean Salinity and Soil Moisture Active Passive L-band radiometer data.This SIT retrieval algorithm was trained using the simulated SIT from the cumulative freezing degree days model during the freeze-up period over five carefully selected regions in the Beaufort,Chukchi,East Siberian,Laptev and Kara seas and utilized the microwave polarization ratio(PR)at incidence angle of 40°.The improvements of the proposed retrieval algorithm include the correction for the sea ice concentration impact,reliable reference SIT data over different representative regions of the Arctic Ocean and the utilization of microwave polarization ratio that is independent of ice temperature.The relationship between the SIT and PR was found to be almost stable across the five selected regions.The SIT retrievals were then compared to other two existing algorithms(i.e.,UH_SIT from the University of Hamburg and UB_SIT from the University of Bremen)and validated against independent SIT data obtained from moored upward looking sonars(ULS)and airborne electromagnetic(EM)induction sensors.The results suggest that the proposed algorithm could achieve comparable accuracies to UH_SIT and UB_SIT with root mean square error(RMSE)being about 0.20 m when validating using ULS SIT data and outperformed the UH_SIT and UB_SIT with RMSE being about 0.21 m when validatng using EM SIT data.The proposed algorithm can be used for thin sea ice thickness(<1.0 m)estimation in the Arctic Ocean and requires less auxiliary data in the SIT retrieval procedure which makes its implementation more practical.

Advances in Research on Soil Moisture by Microwave Remote Sensing in China

Soil moisture is an important factor in global hydrologic circulation and plays a significant role in the research of hydrology, climatology, and agriculture. Microwave remote sensing is less limited by climate and time, and can measure in large scale. With these characteristics, this technique becomes an effective tool to measure soil moisture. Since the 1980s, Chinese researchers have investigated the soil moisture using microwave instruments. The active re- mote sensors are characteristic of high spatial resolution, thus with launch of a series of satellites, active microwave remote sensing of soil moisture will be emphasized. The passive microwave remote sensing of soil moisture has a long research history, and its retrieval algorithms were developed well, so it is an important tool to retrieve large scale moisture information from satellite data in the future.

Coseismic landslide sediment increased by the“9.5”Luding earthquake,Sichuan,China

The"9.5"Luding earthquake(Ms 6.8),which occurred on September 5,2022,has triggered thousands of landslides,and caused coseismic landslide sediment in the mountain basin to increase significantly.After the Luding earthquake,landslide sediment may continue to divert to channels,and increase the activity of debris flows.Importantly,the formation of debris flows can pose a major threat to infrastructure,lives and property.To better understand the landslide sediment that increased by the"9.5"Luding earthquake and its impact on the activity of debris flows,we mapped the coseismic landslide database using satellite images.A total of9142 landslides with an area of 49.51 km^(2),covering4.81%of the whole basin,were triggered by the Luding earthquake.The coseismic landslides induced by this earthquake are dominated by shallow landslides and are densely distributed in the combined zone of the Xianshuihe fault and the Daduhe fault.Approximately 333.31×10^(6)m^(3)(error:111.43×10^(6)m^(3)/-70.73×10^(6)m^(3))of coseismic landslide sediments were induced by the earthquake in the epicenter,and the landslide materials were concentrated downstream of the basins.In addition.more than 13986.45×10^(4)m^(3)(error:4675.67×10^(4)m^(3)/-2967.92×10^(4)m^(3))of landslide sediment may supply for debris flow occurrence.Simultaneously,the small basins that are distributed near Moxi,Detuo and the junction of the Xianshuihe fault and Daduhe fault are more susceptible to debris flows when rainstorms hit these regions.Therefore,prevention and mitigation measures,early warning,and land use planning should be adopted in advance in these regions.However,from the perspectives of landslide scale and the degree of landslide-channel coupling,the activity or active intensity of debris flows in the Luding earthquake area may be lower than that in the epicenter area of the 2008 Wenchuan earthquake.

改进YOLOv4的遥感图像目标检测算法

为有效解决遥感图像目标检测算法在复杂背景下的检测效果不佳的问题,提出一种改进YOLOv4的目标检测算法。设计一种跨阶段残差结构,替换原主干网络的简单残差结构,降低模型参数量和计算负担;引入CBAM注意力机制,加强CSP模块间有效特征交互;使用跨阶段分层卷积模块重构特征融合阶段对深层特征图的处理方式,防止网络退化和梯度消失;采用Mish激活函数,增强融合网络对非线性特征的提取能力。在RSOD、DIOR数据集上的实验结果表明,改进YOLOv4算法的测试mAP相比原YOLOv4算法分别高出4.5%、7.3%,其检测速度分别达到48 fps、45 fps,在保证实时性的同时检测精度有较大提升。

多级对比学习下的弱监督高分遥感影像城市固废堆场提取

城市固体废物是城市化进程中的重要污染源,对城市生态环境和公共健康造成了巨大危害。高分影像固废堆场智能解译是实现自动排查,提升监测效率的核心和关键技术。基于深度学习的固废堆场自动提取方法严重依赖于获取成本高、制作难度大的高质量像素级标注。为此,本文提出使用更易获取的影像级标注,利用影像自监督学习实现像素级固废堆场提取。围绕固废堆场的影像特征,本文方法在尺度对比约束下综合像素、影像两个层次的对比学习方法,对固废堆场的类别激活图细化和完善,并基于此生成高质量的固废堆场伪像素级标注,用于训练固废堆场提取模型。试验结果表明,本文方法在固废堆场提取的F 1值和IoU分数方面分别达到了71.58%和55.74%,显著优于所有对比方法。这说明利用多级对比学习的弱监督方法能够获得更加完整且准确的类别激活图,从而取得更高的固废堆场提取精度。

基于微波遥感的土壤水分反演估算研究进展

土壤水分是地表和大气水热过程交换的重要纽带,对于农业生产、生态规划、水资源管理等具有十分重要的意义。微波遥感具有基本不受天气条件影响,具有较好探测植被覆盖下的土壤信息和土壤水分变化趋势等优势,成为目前遥感精确反演土壤水分的热点。本文整理了现有全球尺度的基于微波遥感的土壤水分产品;分析比较了土壤水分反演中主动微波遥感、被动微波遥感、主被动微波协同技术的原理、特点、适用范围和关键技术进展:主动微波遥感和被动微波遥感的优势分别在于高空间分辨率和高时间分辨率,高空间分辨率可以很好捕捉地表细微的空间信息特征,但囿于土壤水分与后向散射系数之间的复杂关系,特别是植被、地表粗糙度等对雷达后向散射系数的干扰,使得反演土壤水分的精度不高,因而根据现实情况选取不同散射模型以及利用多源数据协同是目前改善精度的研究热点。而高时间分辨率可以实现全球及大尺度下的土壤水分监测,但是很难满足小尺度或者小区域范围的实际研究需求,为了能使实测数据在空间上得以较好匹配,提出多种降尺度方法。结合以上两种微波遥感方式的优劣,依托更为丰富的数据源、相对成熟的观测技术来对两者进行融合以提取更多的水分信息,以提升反演精度或者获得长时间序列数据。在目前的方法中,土壤水分反演在小尺度下表现出良好的性能,但在全球尺度上会出现数据缺失、适用性不强、反演精度不高以及反演过程过于复杂等诸多问题,可以借助多种观测方式(多极化、多角度、多波段)、多时相重复观测、在原有模型上引入新的算法以及数据同化等方面着手进行改进,同时全球卫星导航系统(Global Navigation Satellite System,GNSS)中长期稳定、高时空分辨率的L波段微波信号在陆面遥感领域的快速发展也为我国北斗卫星导航系统(BeiDou Navigation Satellite System,BDS)的发展提供了借鉴,展现出在土壤水分反演方面的巨大潜力。

基于多特征融合神经网络的遥感影像推荐方法

针对目前遥感用户获取遥感影像时采用的人工查询和订购方式较为低效的问题,提出了一种基于多特征融合神经网络的遥感影像个性化推荐框架。设计遥感影像语义体系实现影像语义提取,结合用户操作记录与职责描述文本构建遥感领域知识图谱。通过嵌入表征方法提取用户与影像的多维度属性特征。设计了一种基于协同过滤的多特征融合神经网络模型,借助神经网络的高维度空间建模能力实现用户与影像多维属性特征的有效融合,达到较好的配对推荐效果。实验结果显示,相较于先前的推荐方法,所提方法的性能获得显著提高,有效提升了遥感数据服务的精准度与时效性。

2018–2020年广西甘蔗10 m分辨率种植分布数据集

作物种植分布的提取是作物监测的基础,及时准确地提取出甘蔗种植分布对甘蔗监测及种植结构调整具有重要意义。本文基于Sentinel-1和Sentinel-2数据,采用主被动遥感协同方法和决策树分类方法,开展了广西甘蔗种植分布提取方法研究,并结合多年调查的地面样本数据进行验证,在此基础上提取了2018–2020年广西甘蔗种植分布。该方法利用遥感数据提取得到的2018年广西甘蔗种植分布总体精度达92%,Kappa系数为0.85;2019年广西甘蔗种植分布总体精度达94%,Kappa系数为0.88;2020年广西甘蔗种植分布总体精度达94%,Kappa系数为0.88。本数据集可作为广西甘蔗时空变化分析的基础数据,也可为广西甘蔗生产管理与种植结构优化调整等提供基础数据支撑。

Remote Sensing Interpretation and Extraction of Structural Information about Active Faults at Hangzhou,China,and Their Surroundings

It is important to explore active faults in urban areas and their surroundings for earth- quake disaster mitigation. Satellite remote sensing techniques can play an important role in such active fault exploration. It can not only reveal the pattern of active faults and active tectonics on a macroscop- ic scale, but also monitor the occurrence, development and rules of temporal-spatial evolution of active faults. In this paper, we use the Hangzhou area as an example to introduce methods of extracting de- tailed active fault information when covered by thick unconsolidated Quaternary sediment, using im- age enhancement and image fusion etc. to improve the definition and precision of satellite images and presenting a three-dimensional （3D） image to illustrate tectono-geomorphic features along the relevant faults. We have also collected aeromagnetic anomaly data, shallow seismic exploration data and dating data, and carried out field surveys to validate the characteristics of active faults based on remote sens- ing images. The results revealed about the faults showed a high consistency with traditional geological knowledge, and demonstrate that it is feasible to explore active faults in a weakly active tectonic area by using satellite remote sensing techniques and contribute to large engineering projects and research on neotectonics.

A reappraisal of active faults in central-east Iran(Kerman province)

Fault lineaments are the main input data in earthquake engineering and seismology studies.This study presents a digitally-based active fault map of the Kerman region in central-east Iran which experienced several devastating earthquakes on poorly exposed and/or not identified active faults.Using Landsat 8 data,we have carried out the image-based procedures of fault mapping,which include applying the contrast stretching technique,the principal component analysis,the color composite technique,the spectral rationing,and creating the false-color composite images.Besides,we have cross-checked the resulting map with the geological maps provided by the Geological Survey of Iran to decrease the associated uncertainties.The resulting map includes 123 fault segments,still,a part of which has been expressed in the previously compiled active-fault maps of Iran.Indeed,the new one is mapping the poorly exposed active faults,so-called secondary faults,which are able to produce strong events.These faults are primarily associated with poorly defined areas that accommodate low levels of seismicity;however,sporadic strong events are likely to occur.It has also been investigated that these kinds of faults are seismogenic and are able to produce destructive events.In total,the outcome of this study can also be jointed with seismic studies for investigating parts of the earthquake activity in central-east Iran,in particular for the fault-based approaches in impending earthquake-resistant buildings.

星载二氧化碳探测研究进展

二氧化碳(CO_(2))是大气中最典型的温室气体,大气中CO_(2)浓度持续增加是全球气温升高的主要原因。温度升高导致自然环境恶劣,各类自然灾害频发,极端天气数量增加,对人类生活环境造成巨大的影响,因此推动绿色低碳发展势在必行。准确监测CO_(2)的排放,了解CO_(2)源和汇的分布,是遏制大气CO_(2)浓度升高、减缓全球变暖的前提条件。因此,准确获取全球范围内高精度时空分辨率的CO_(2)观测数据十分必要。卫星遥感可以有效观测CO_(2)的全球分布,获取大范围高精度数据。本文旨在对目前用于大气CO_(2)探测的卫星载荷发展状况进行综述,包括基于被动遥感CO_(2)的卫星现状和数据产品、基于主动遥感CO_(2)的卫星的研制情况和目前的主要工作进展。

融合主被动遥感影像的冬小麦种植面积提取研究

为了快速、准确地获取作物分布信息,探索使用主动遥感影像(Sentinel-1A)和被动遥感影像(Sentinel-2)提取冬小麦空间分布的可行性。首先,根据冬小麦的物候特征,合成冬小麦全生育期的Sentinel-1A影像;并依据各类地物的NDVI(归一化植被指数)时序曲线合成一期高质量的冬小麦越冬后Sentinel-2影像。其次,设计Sentinel-1A影像、Sentinl-2影像和融合Sentinel-1A与Sentinl-2主被动遥感影像3种分类方案,然后在Google Earth Engine(GEE)云平台上基于随机森林算法对冬小麦进行分类。结果表明,基于全生育期Sentinel-1A影像的冬小麦用户精度和生产者精度分别为83.15%和86.44%,提取结果中存在较多的“椒盐”噪声;基于冬小麦越冬后Sentinl-2影像的冬小麦用户精度和生产者精度分别为87.98%和84.75%,提取精度较使用全生育期Sentinel-1A影像有所提高,但分类结果受“异物同谱”的影响,产生许多错分;融合主被动遥感影像的冬小麦用户精度和生产者精度分别为96.57%和95.48%,相较于仅使用单一数据源,冬小麦分类精度有不同程度的提升。

主被动光学卫星遥感数据融合浅海水深测量

针对双介质立体摄影测量中水面高程不确定性带来的测深误差,本文提出了主被动光学卫星遥感数据融合浅海水深测量的技术路线,探讨了主动ICESat-2激光测高数据与被动WorldView-3立体像对数据融合的测深模型。首先通过ICESat-2激光测高数据提取海面高程作为WorldView-3立体像对数据水面高程纠正的起算位置,然后利用区域网平差的二次多项式加入适当的水深改正量。在我国南海岛礁开展了主被动光学卫星遥感数据融合进行浅海水深测量试验。结果表明,主被动融合测深模型的均方根误差(RMSE)小于0.70 m,实现了高精度、无实测控制点的浅海水深测量。

大气环境监测卫星方案设计与技术特点

大气环境监测卫星(DQ-1)是国家民用空间基础设施中长期发展规划中的科研卫星,其运行于705 km高度的太阳同步轨道,装载大气探测激光雷达(ACDL)、高精度偏振扫描仪(POSP)、多角度偏振成像仪(DPC)、紫外高光谱大气成分探测仪(EMI)及宽幅成像光谱仪(WSI)等5台遥感仪器,通过主动激光与被动高光谱、多光谱、多角度、偏振等手段结合,实现对大气CO_(2)、细颗粒物、污染气体、云和气溶胶等要素,以及对大气环境、水环境和生态环境等进行大范围、连续、动态、全天时的综合监测;卫星CO_(2)柱浓度探测精度优于1×10^(-6),为国际最高。卫星入轨后各分系统工作正常,在轨测试结果均满足要求,能够进一步提升我国大气环境综合监测、全球气候变化和农作物估产、农业灾害等应对能力,推动生态环境、气象、农业农村等领域遥感应用。本文概述了DQ-1卫星的总体设计方案,总结了卫星的主要技术特点及创新点,介绍了卫星系统在轨测试情况,并对卫星的应用前景进行了展望。

基于改进D-Linknet的高分遥感影像道路提取方法

基于D-Linknet进行图像分割时,在特征提取阶段忽略了不同级别的特征图中存在的特征重要程度不同,丢失了大量的细节信息,致使分割效果不佳。针对该问题,提出了一种基于D-Linknet与注意力机制相结合的图像语义分割算法,并将上采样网络的所有子区块的输出进行聚类后再恢复至原始图片维度,以提高上采样时子区块的利用率;通过采用最大特征映射(Max-feature-map,MFM)作为激活函数,解决部分稀疏道路提取不成功的问题。实验结果表明,提出的网络精准率达到85.1%,平均交并比达到85.6%,F1-score达到84.6%;相比于同类网络,评价指标也均处于前列。

基于YOLOv5的光学遥感图像舰船目标检测算法

针对YOLO(you only look once)v5算法在应用于光学遥感图像舰船目标检测任务时所面临的小目标误检率、漏检率较高的情况,提出一种基于YOLOv5改进的光学遥感图像舰船目标检测方法。首先对路径聚合网络结构进行改进,设计语义信息增强模块提取更富语义信息的浅层特征,增强对小目标特征的表达能力;然后使用Swish函数作为激活函数,提高网络对数据非线性特征的表征能力,加快模型的收敛速度;最后针对舰船目标的尺寸特点优化检测端结构,移除大目标检测头以减少推理计算量。测试集上的实验表明,该方法相较改进前将检测精度提高了5.2%且推理时间有所减少,在保证检测实时性的同时增强了模型的小目标辨别能力。