High-resolution Remote Sensing Image Segmentation Using Minimum Spanning Tree Tessellation and RHMRF-FCM Algorithm

It is proposed a high resolution remote sensing image segmentation method which combines static minimum spanning tree(MST)tessellation considering shape information and the RHMRF-FCM algorithm.It solves the problems in the traditional pixel-based HMRF-FCM algorithm in which poor noise resistance and low precision segmentation in a complex boundary exist.By using the MST model and shape information,the object boundary and geometrical noise can be expressed and reduced respectively.Firstly,the static MST tessellation is employed for dividing the image domain into some sub-regions corresponding to the components of homogeneous regions needed to be segmented.Secondly,based on the tessellation results,the RHMRF model is built,and regulation terms considering the KL information and the information entropy are introduced into the FCM objective function.Finally,the partial differential method and Lagrange function are employed to calculate the parameters of the fuzzy objective function for obtaining the global optimal segmentation results.To verify the robustness and effectiveness of the proposed algorithm,the experiments are carried out with WorldView-3(WV-3)high resolution image.The results from proposed method with different parameters and comparing methods(multi-resolution method and watershed segmentation method in eCognition software)are analyzed qualitatively and quantitatively.

Assessment of the State of Forests Based on Joint Statistical Processing of Sentinel-2B Remote Sensing Data and the Data from Network of Ground-Based ICP-Forests Sample Plots

The research was carried out on the territory of the Karelian Isthmus of the Leningrad Region using Sentinel-2B images and data from a network of ground sample plots. The ground sample plots are located in the studied territory mainly in a regular manner, laid and surveyed according to the ICP-Forests methodology with some additions. The total area of the sample plots is a small part of the entire study area. One of the objectives of the study was to determine the possibility of using the k-NN (nearest neighbor method) to assess the state of forests throughout the whole studied territory by joint statistical processing of data from ground sample plots and Sentinel-2B imagery. The data of the ground-based sample plots were divided into 2 equal parts, one for the application of the k-NN method, the second for checking the results of the method application. The systematic error in determining the mean damage class of the tree stands on sample plots by the k-NN method turned out to be zero, the random error is equal to one point. These results offer a possibility to determine the state of the forest in the entire study area. The second objective of the study was to examine the possibility of using the short-wave vegetation index (SWVI) to assess the state of forests. As a result, a close statistically reliable dependence of the average score of the state of plantations and the value of the SWVI index was established, which makes it possible to use the established relationship to determine the state of forests throughout the studied territory. The joint use and statistical processing of remotely sensed data and ground-based test areas by the two studied methods make it possible to assess the state of forests throughout the large studied area within the image. The results obtained can be used to monitor the state of forests in large areas and design appropriate forestry protective measures.

Sentinel-2遥感影像在盘锦水稻米质监测中的应用研究

本研究基于水稻孕穗期、抽穗期、灌浆期和成熟期4个生育期的Sentinel-2遥感数据,分析各生育期内卫星遥感光谱参数与稻米品质指标的关系,建立基于各生育期卫星光谱信息的水稻品质指标预测模型。将5种稻米品质指标分别与4个生育期内的光谱参数进行皮尔逊相关性分析,结果表明,5项品质指标在4个生育期内均与光谱参数有不同程度相关性。然后筛选出相关性效果显著的光谱参数,用于建立各品质指标的预测方程,建模结果表明,基于卫星遥感光谱信息解释率由大到小的稻米品质指标依次是精米率>长宽比>蛋白质含量>直链淀粉含量>糙米率;卫星遥感光谱反演稻米各品质指标所在的最佳生育期不同,糙米率和精米率的最佳生育期为抽穗期,其建模决定系数(Coefficient of Determination,R^(2))分别为0.461和0.893;长宽比的最佳生育期为成熟期,R^(2)为0.878;直链淀粉含量和蛋白质含量的最佳生育期为灌浆期,R^(2)分别为0.646和0.647;基于卫星遥感光谱信息的稻米品质模型验证效果较好,解释率为51%~74%。可见,利用卫星遥感技术能够实现大范围水稻品质指标定量监测与评估。

A Remote Sensing Image Semantic Segmentation Method by Combining Deformable Convolution with Conditional Random Fields

Currently,deep convolutional neural networks have made great progress in the field of semantic segmentation.Because of the fixed convolution kernel geometry,standard convolution neural networks have been limited the ability to simulate geometric transformations.Therefore,a deformable convolution is introduced to enhance the adaptability of convolutional networks to spatial transformation.Considering that the deep convolutional neural networks cannot adequately segment the local objects at the output layer due to using the pooling layers in neural network architecture.To overcome this shortcoming,the rough prediction segmentation results of the neural network output layer will be processed by fully connected conditional random fields to improve the ability of image segmentation.The proposed method can easily be trained by end-to-end using standard backpropagation algorithms.Finally,the proposed method is tested on the ISPRS dataset.The results show that the proposed method can effectively overcome the influence of the complex structure of the segmentation object and obtain state-of-the-art accuracy on the ISPRS Vaihingen 2D semantic labeling dataset.

基于Sentinel-2卫星影像的海南西岛珊瑚礁识别和变化分析

珊瑚礁是海洋中最重要的生态系统之一,近年来在全球气候变化和人为干扰加剧的影响下,我国南海珊瑚礁总体处于快速退化状态。以海南西岛珊瑚礁为例,基于Sentinel-2系列卫星10 m空间分辨率影像,利用面向对象分类法(object-based image analysis,OBIA)对2017年12月~2018年3月和2021年12月两个时期的海南西岛珊瑚礁底质进行了识别分类,并进行珊瑚礁面积变化分析。将2021年12月的分类结果与现场调查数据进行对比验证,总分类精度和Kappa系数分别为83.3%和0.71。对比两个时期珊瑚礁底质分类结果表明,西岛西侧珊瑚礁覆盖面积未出现明显变化,东侧珊瑚礁显示恢复趋势。本文研究表明,10 m地面分辨率卫星系列影像和面向对象的阈值分类方法可以对海南西岛珊瑚礁进行较为准确的识别和变化分析,监测结果可为海南岛沿岸西岛等小型岛礁珊瑚保护及修复提供参考。

基于多时相Sentinel-2卫星影像的冬小麦面积提取

及时准确地提取冬小麦种植信息,对开展冬小麦农情遥感监测具有重要的意义.以杭州市余杭区冬小麦越冬期(2021-12-04)、扬花期(2022-04-08)和乳熟期(2022-05-03)Sentinel-2遥感影像为数据源,分别采用最大似然法、支持向量机、归一化差值植被指数(normalized difference vegetation index,NDVI)相加和相减合成运算提取冬小麦种植面积.结合冬小麦地面调查数据与实测种植面积,对不同方法的提取结果进行精度评价.结果显示,利用越冬期影像NDVI阈值将常绿植被区(茶园、林地)掩膜处理,对非常绿植被区(建筑、水体、冬小麦)扬花期与乳熟期影像NDVI值进行和值运算,是提取余杭区冬小麦种植面积的最佳方法,面积精度为91.96%,说明基于多时相遥感影像结合植被物候特征与典型地物类型,能够实现冬小麦种植面积的高精度提取.

基于Sentinel-2影像的果树提取方法及其空间分析研究——以甘肃省平凉市为例

利用遥感技术对果园进行快速监测,准确掌握苹果园地面积与空间种植分布状况,有助于促进当地经济的发展。目前针对丘陵区果园提取的研究较少,相关方法的有效性和可靠性仍然存在问题。以甘肃省平凉市为研究区域,采用NDVI,RVI,EVI,SIPI,LSWI,NDWI等指标对输入数据进行增强,通过基于数据增强的梯度提升树算法提取研究区苹果种植面积。为验证该方法的有效性,引入最小距离法、CART决策树法、支持向量机法和随机森林4种机器学习算法进行对比分析,结果表明,梯度提升树算法分类精度最高,总体分类精度(Overall Accuracy,OA)达到89.3%,Kappa系数为0.77,分类效果及一致性均最佳。此外,采用基于数据增强的梯度提升树法分别对2019—2023年的苹果园进行提取,获得平凉市苹果园种植变化情况,各区县苹果园种植面积除泾川县外整体呈现上升趋势,泾川县和静宁县种植面积最大,其次为庄浪县、灵台县和崆峒区,最小的为崇信县和华亭市。

A landslide extraction method of channel attention mechanismU-Net network based on Sentinel-2A remote sensing images

Accurate landslide extraction is significant for landslide disaster prevention and control.Remote sensing images have been widely used in landslide investigation,and landslide extraction methods based on deep learning combined with remote sensing images(such as U-Net)have received a lot of attention.However,because of the variable shape and texture features of landslides in remote sensing images,the rich spectral features,and the complexity of their surrounding features,landslide extraction using U-Net can lead to problems such as false detection and missed detection.Therefore,this study introduces the channel attention mechanism called the squeeze-and-excitation network(SENet)in the feature fusion part of U-Net;the study also constructs an attention U-Net landside extraction model combining SENet and U-Net,and uses Sentinel-2A remote sensing images for model training and validation.The extraction results are evaluated through different evaluation metrics and compared with those of two models:U-Net and U-Net Backbone(U-Net Without Skip Connection).The results show that proposed the model can effectively extract landslides based on Sentinel-2A remote sensing images with an F1 value of 87.94%,which is about 2%and 3%higher than U-Net and U-Net Backbone,respectively,with less false detection and more accurate extraction results.

Monitoring of vegetation coverage based on high-resolution images

Measurement of vegetation coverage on a small scale is the foundation for the monitoring of changes in vegetation coverage and of the inversion model of monitoring vegetation coverage on a large scale by remote sensing. Using the object-oriented analytical software, Definiens Professional 5, a new method for calculating vegetation coverage based on high-resolution images （aerial photographs or near-surface photography） is proposed. Our research supplies references to remote sensing measurements of vegetation coverage on a small scale and accurate fundamental data for the inversion model of vegetation coverage on a large and intermediate scale to improve the accuracy of remote sensing monitoring of changes in vegetation coverage.

RepDDNet:a fast and accurate deforestation detection model with high-resolution remote sensing image

Forest is the largest carbon reservoir and carbon absorber on earth.Thus,mapping forest cover change accurately is of great significance to achieving the global carbon neutrality goal.Accurate forest change information could be acquired by deep learning methods using high-resolution remote sensing images.However,deforestation detection based on deep learning on a large-scale region with high-resolution images required huge computational resources.Therefore,there was an urgent need for a fast and accurate deforestation detection model.In this study,we proposed an interesting but effective re-parameterization deforestation detection model,named RepDDNet.Unlike other existing models designed for deforestation detection,the main feature of RepDDNet was its decoupling feature,which means that it allowed the multi-branch structure in the training stages to be converted into a plain structure in the inference stage,thus the computation efficiency can be significantly improved in the inference stage while maintaining the accuracy unchanged.A large-scale experiment was carried out in Ankang city with 2-meter high-resolution remote sensing images(the total area of it was over 20,000 square kilometers),and the result indicated that the model computation efficiency could be improved by nearly 30%compared with the model without re-parameterization.Additionally,compared with other lightweight models,RepDDNet also displayed a trade-off between accuracy and computation efficiency.

Monitoring the green evolution of vernacular buildings based on deep learning and multi-temporal remote sensing images

The increasingly mature computer vision(CV)technology represented by convolutional neural networks(CNN)and available high-resolution remote sensing images(HR-RSIs)provide opportunities to accurately measure the evolution of natural and artificial environments on Earth at a large scale.Based on the advanced CNN method high-resolution net(HRNet)and multi-temporal HR-RSIs,a framework is proposed for monitoring a green evolution of courtyard buildings characterized by their courtyards being roofed(CBR).The proposed framework consists of an expert module focusing on scenes analysis,a CV module for automatic detection,an evaluation module containing thresholds,and an output module for data analysis.Based on this,the changes in the adoption of different CBR technologies(CBRTs),including light-translucent CBRTs(LT-CBRTs)and non-lighttranslucent CBRTs(NLT-CBRTs),in 24 villages in southern Hebei were identified from 2007 to 2021.The evolution of CBRTs was featured as an inverse S-curve,and differences were found in their evolution stage,adoption ratio,and development speed for different villages.LT-CBRTs are the dominant type but are being replaced and surpassed by NLT-CBRTs in some villages,characterizing different preferences for the technology type of villages.The proposed research framework provides a reference for the evolution monitoring of vernacular buildings,and the identified evolution laws enable to trace and predict the adoption of different CBRTs in a particular village.This work lays a foundation for future exploration of the occurrence and development mechanism of the CBR phenomenon and provides an important reference for the optimization and promotion of CBRTs.

基于GF-2遥感影像和改进后PSPNet模型的丘陵地区耕地图斑提取方法

针对丘陵地区耕地地块具有边界模糊、覆盖物种类多样、大小和空间位置分布不规则等特点,传统分类方法难以快速准确提取耕地信息的问题,本文以四川省金堂县竹篙镇和高板镇为研究区域,利用高分二号卫星影像和改进后的PSPNet语义分割网络模型进行耕地图斑提取研究。在模型训练中,引入CBAM注意力模块以提高整个网络的特征提取和表达能力,采用余弦退火学习率以加快模型的收敛速度。结果表明,改进后的PSPNet模型在丘陵地区耕地提取精度方面取得了显著提高,耕地识别精度达到了95.69%,比标准PSPNet模型提高了1.07%,比Unet++,DeepLabv3+和支持向量机方法方法提高了1.32%,1.75%和6.33%。基于改进后的PSPNet模型具有更强的特征提取和表达能力,可以更准确地提取丘陵地区的耕地信息,为农业决策提供更准确的数据支持,促进农业智能化和精准化,提高农作物产量和质量,推动农业现代化进程。

U^(2)-Net在建筑物提取中的边缘精度分析

随着城市化的进程和遥感科学技术的发展,在高分辨遥感影像中进行建筑物提取一直是摄影测量与遥感领域的一个热点研究主题。针对遥感影像中提取建筑物存在边缘模糊的问题,本文运用U^(2)-Net网络算法提取建筑物,并与lr-aspp、fcn、deeplab_v3三种网络算法分别进行了建筑物提取对比实验;结果表明U^(2)-Net网络,在不损失预测精度的情况下,耗时较短,且准确率可提升至97.478%,可较好地解决建筑物提取中的边缘模糊问题。

一种GF-2全色多光谱影像融合方法

高分2号卫星影像的应用日益多样化,如何获取高质量融合影像成为需要研究的重要问题。为了从原始的全色影像和多光谱影像中获取高空间分辨率和高光谱分辨率的影像,结合方向信息和脉冲耦合神经网络对非下采样轮廓波变换算法进行改进。以甘肃省兰州市的GF-2全色影像和多光谱影像作为数据源,提出了一种结合PCNN和NSCT的遥感影像融合方法,通过定性评估和定量评估,与IHS方法、BT方法、PCA方法和GS方法相比,改进后的方法在改善空间细节和保留光谱信息方面具有更好的效果。

基于Sentinel-2遥感影像的玉米冠层叶面积指数反演

叶面积指数是描述玉米冠层结构的重要参数之一,决定玉米冠层的光合作用、呼吸作用、蒸腾和碳循环等生物物理过程,因此精确反演叶面积指数对玉米长势监测具有重要意义。以河北省保定市的涿州市、高碑店市、定兴县为研究区,利用Sentinel-2遥感影像和LAI-2000地面同步实测数据进行玉米冠层叶面积指数反演,使用归一化差异光谱指数和比值型光谱指数两类指数,构建了单变量和多变量玉米冠层叶面积指数反演模型,通过决定系数(R2)和均方根误差(RMSE)筛选出最佳模型。研究结果表明,由NDSI(783,705)构建的单变量模型为最优反演模型,其决定系数为0.534 2,均方根误差为0.288 5。因此,基于Sentinel-2遥感影像利用植被指数反演玉米冠层叶面积指数的方法可作为判断玉米长势状况的初步判断依据。

基于Sentinel-2影像的围海养殖信息提取

为了能够利用遥感图像快速准确地提取围海养殖矢量信息,本文选取养殖水体、堤坝及育苗室等交错分布的海参围海养殖区域作为研究区域,根据研究区域Sentinel-2遥感影像的光谱特征,选用归一化差异水体指数(Normalized Difference Water Index,NDWI)、改进归一化差异水体指数(Modified Normalized Difference Water Index,MNDWI)和增强水体指数(Enhanced Water Index,EWI)三类水体指数,分别进行提取实验,利用同时期高空间分辨率的高分二号卫星(GF-2)影像作为参考,验证不同方法的提取精度,精度评价结果表明:相较MNDWI和EWI两类水体指数,NDWI的分类精度更高,且利用NDWI提取研究区域的围海养殖信息的效果更好,所以该方法可在养殖区域的动态监测和规划管理中发挥数据支撑作用。

Big geochemical data through remote sensing for dynamic mineral resource monitoring in tailing storage facilities

Evolution in geoscientific data provides the mineral industry with new opportunities.A direction of geochemical data generation evolution is towards big data to meet the demands of data-driven usage scenarios that rely on data velocity.This direction is more significant where traditional geochemical data are not ideal,which is the case for evaluating unconventional resources,such as tailing storage facilities(TSFs),because they are not static due to sedimentation,compaction and changes associated with hydrospheric and lithospheric processes(e.g.,erosion,saltation and mobility of chemical constituents).In this paper,we generate big secondary geochemical data derived from Sentinel-2 satellite-remote sensing data to showcase the benefits of big geochemical data using TSFs from the Witwatersrand Basin(South Africa).Using spatially fused remote sensing and legacy geochemical data on the Dump 20 TSF,we trained a machine learning model to predict in-situ gold grades.Subsequently,we deployed the model to the Lindum TSF,which is 3 km away,over a period of a few years(2015-2019).We were able to visualize and analyze the temporal variation in the spatial distributions of the gold grade of the Lindum TSF.Additionally,we were able to infer extraction sequencing(to the resolution of the data),acid mine drainage formation and seasonal migration.These findings suggest that dynamic mineral resource models and live geochemical monitoring(e.g.,of elemental mobility and structural changes)are possible without additional physical sampling.

基于U 2-Net深度学习模型的沿海水产养殖塘遥感信息提取

针对近海沿岸复杂地理环境中“同谱异物”效应导致传统方法提取水产养殖塘边界模糊、精度较低的问题,提出了基于U 2-Net深度学习模型的沿海水产养殖塘遥感信息提取方法。首先,对遥感影像进行预处理,选择合适的波段组合方式以区分养殖塘和其他地物;其次,通过目视解译进行样本制作;然后,利用U 2-Net深度学习模型训练并提取沿岸养殖塘;最后,利用局部最佳法确定养殖塘范围。实验结果表明,该方法平均总体精度达到95.50%,平均Kappa系数、召回率和F值分别为0.91,91.45%和91.01%;在养殖塘个数及面积评价方面,提取出养殖塘区19块,共计9.79 km^(2),区块数和面积的平均准确度分别为94.06%和93.18%。本研究能够快速、准确地开展海岸带区域养殖塘制图,能够为海洋资源管理和可持续发展提供技术支持。

基于Sentinel-2影像的雄安新区2016—2019年土地利用分析

雄安新区是国家层面打造的又一个具有重要战略意义的新区,及时准确掌握该地区的土地利用详情具有重要意义。本文利用10 m分辨率的Sentinel-2影像对雄安新区2016—2019年的土地利用进行分类,进而分析该地区的土地利用时空演变。共测试了决策树、随机森林和支持向量机3种分类器,进而获得最高精度的土地分类结果图;同时,利用随机森林的特征排序功能分析了不同特征的重要性。结果表明,雄安新区的耕地、林地、水生植物面积总体均呈显著减少趋势,建设用地面积变化最为显著,表明雄安新区正在进行中、快速的城市化发展。本研究得到的10 m分辨率土地利用专题图和分析结果对于雄安新区的及时监测与规划有着重要参考意义。

基于Sentinel-2与机载激光雷达数据的误差变量联立方程组森林蓄积量反演研究

【目的】研究通过提取Sentinel-2中的特征变量与机载激光雷达(Light detection and ranging,LiDAR)中的冠层高度模型(Canopy height model,CHM),探索使用误差变量联立方程组反演森林蓄积量制图的新方法。【方法】以广西壮族自治区国有高峰林场的界牌与东升分场为研究区,机载LiDAR和Sentinel-2影像为数据源,利用皮尔森相关系数与方差膨胀因子(Variance inflation factor,VIF)结合线性逐步回归进行遥感特征变量筛选。通过VIF判断和线性逐步回归保留后的遥感特征变量与LiDAR提取的CHM,分别选用普通回归模型(多元线性逐步回归与Logistic模型)、误差变量联立方程组、随机森林(Random forest,RF)、kNN(k-Nearest Neighbor,kNN)4种反演方法开展森林蓄积量反演,并利用地面实测数据对反演结果进行验证。【结果】1)在普通回归模型中,Logistic模型的反演精度(RRMSE=30.41%)优于MLR模型的反演精度(RRMSE=30.53%);2)在误差变量联立方程组反演方法中,MLR-Logistic联立模型精度(RRMSE=29.29%)优于Logistic-Logistic、MLR-MLR与Logistic-MLR联立模型(RRMSE分别为29.40%、29.60%与29.66%);3)在4种反演方法中,误差变量联立方程组反演结果精度最高(R2=0.60),显著优于普通回归模型方法、随机森林与kNN反演方法(R2分别为0.56、0.39与0.28)。【结论】误差变量联立方程组反演方法更适用于森林蓄积量遥感估测,其反演精度最高,且获得的蓄积量空间连续分布结果与实际接近,制图效果最好,表明误差变量联立方程组反演森林蓄积量制图方法是可行的。