Estimating fractional cover of crop,crop residue,and soil in cropland using broadband remote sensing data and machine learning     

《International Journal of Applied Earth Observation and Geoinformation》2020

The fractional vegetation cover (FVC), crop residue cover (CRC), and bare soil (BS) are three important parameters in vegetation–soil ecosystems, and their correct and timely estimation can improve crop monitoring and environmental monitoring. The triangular space method uses one CRC index and one vegetation index to create a triangular space in which the three vertices represent pure vegetation, crop residue, and bare soil. Subsequently, the CRC, FVC, and BS of mixed remote sensing pixels can be distinguished by their spatial locations in the triangular space. However, soil moisture and crop-residue moisture (SM-CRM) significantly reduce the performance of broadband remote sensing CRC indices and can thus decrease the accuracy of the remote estimation and mapping of CRC, FVC, and BS. This study evaluated the use of broadband remote sensing, the triangular space method, and the random forest (RF) technique to estimate and map the FVC, CRC, and BS of cropland in which SM-CRM changes dramatically. A spectral dataset was obtained using: (1) from a field-based experiment with a field spectrometer; and (2) from a laboratory-based simulation that included four distinct soil types, three types of crop residue (winter-wheat, maize, and rice), one crop (winter wheat), and varying SM-CRM. We trained an RF model [designated the broadband crop-residue index from random forest (CRRF)] that can magnify spectral features of crop residue and soil by using the broadband remote sensing angle indices as input, and uses a moisture-resistant hyperspectral index as the target. The effects of moisture on crop residue and soil were minimized by using the broadband CRRF. Then, the CRRF-NDVI triangular space method was used to estimate and map CRC, FVC, and BS. Our method was validated by using both laboratory- and field-based experiments and Sentinel-2 broadband remote-sensing images. Our results indicate that the CRRF-NDVI triangular space method can reduce the effect of moisture on the broadband remote-sensing of CRC, and may also help to obtain laboratory and field CRC, FVC, and BS. Thus, the proposed method has great potential for application to croplands in which the SM-CRM content changes dramatically.  相似文献   

高分三号SAR影像双阈值变化检测   总被引：1，自引：1，他引：0  

崔斌  张永红  闫利  魏钜杰 《遥感学报》2020,24(1):1-10

双阈值合成孔径雷达SAR(Synthetic Aperture Radar)变化检测算法具有在发现变化区域的同时还能确定地表发生后向散射变化类型的优点。针对广义高斯双阈值最小误差法D-GKIT(Dual Generalized Kittler and Illingworth Thresholding)在进行阈值选取时直方图中不同类别像素灰度级重叠严重时,分割结果容易在尖峰单侧选取出双阈值而导致无法正确分割差异图的问题,本文提出一种结合归一化最大类间方差和广义高斯最小误差法GKIT(Generalized Kittler and Illingworth Thresholding)的双阈值SAR变化检测方法。首先,提出以归一化最大类间方差值作为灰度级重叠程度的判别参数,确定阈值的选取顺序及两个候选区间;然后,利用GKIT在候选区间内进行分割,获取单侧阈值及非变化类拟合函数;最后,提出利用非变化类拟合函数更新后的直方图作为另一侧阈值选取基础进行分割,得到对应分割阈值。以宁波地区高分三号(GF-3)SAR卫星影像作为试验研究数据,结果表明:本文方法能较好地解决灰度级重叠时D-GKIT无法进行正确分割的问题,具有良好的变化检测效果和更强的鲁棒性且达到了利用研究区数据验证利用GF-3号SAR卫星影像进行变化检测研究可行性的目的。  相似文献   

波段选择协同表达高光谱异常探测算法     

朱德辉  杜博  张良培 《遥感学报》2020,24(4):427-438

高光谱遥感影像具有光谱分辨率极高的特点,承载了大量可区分不同类型地物的诊断性光谱信息以及区分亚类相似地物之间细微差别的光谱信息,在目标探测领域具有独特的优势。与此同时,高光谱遥感影像也带来了数据维数高、邻近波段之间存在大量冗余信息的问题,高维度的数据结构往往使得高光谱影像异常目标类和背景类之间的可分性降低。为了缓解上述问题,本文提出了一种基于波段选择的协同表达高光谱异常探测算法。首先,使用最优聚类框架对高光谱波段进行选择,获得一组波段子集来表示原有的全部波段,使得高光谱影像异常目标类与背景类之间的可分性增强。然后使用协同表达对影像上的像元进行重建,由于异常目标类和背景类之间的可分性增强,对异常目标像元进行协同表达时将会得到更大的残差,异常目标像元的输出值增大,可以更好地实现异常目标和背景类的分离。本文使用了3组高光谱影像数据进行异常目标探测实验,实验结果表明,该方法与其他现有高光谱异常目标探测算法对比,曲线下面积AUC(Area Under Curve)值更高,可以更好地实现异常目标与背景分离,能够更有效地对高光谱影像进行异常目标探测。  相似文献   

姿态修正技术在黄河河口测验中的应用     

宋中华  田慧  王静 《测绘通报》2020,(11):120-123

为了解决黄河三角洲附近海区测验中测船姿态对单波束数字测深数据的影响，本文利用姿态传感器对测船的姿态进行了有效改正。试验结果表明，姿态修正技术对数字测深仪瞬时水深进行改正后，其测深结果与传统的人工水深曲线改正后成果比较，中误差为0.089 m，试验段面0 m线以下面积相差不超过0.22%，符合国家规范要求，提高了海区测验中测深成果的精度，为数字技术在海区测量的应用提供了技术保证。  相似文献   

基于深度学习理论的无人机影像分类     

阳成 《北京测绘》2020,(4):481-484

针对无人机影像深度学习分类方法缺乏现状,本文利用深度学习理论卷积神经网络方法对无人机影像进行了分类。该法首先抽取无人机影像作为训练集和检验集,然后建立一个2个卷积层-池化层的卷积神经网络模型进行深度学习,通过设定参数并运行模型实现无人机影像分类。实验表明,本文提出的方法可完成较复杂地区无人机影像分类,其分类精度与支持向量机方法相当,为无人机遥感影像分类提供了一个崭新的技术视点。  相似文献   

山地叶面积指数反演理论、方法与研究进展   总被引：2，自引：0，他引：2  

江海英  贾坤  赵祥  魏香琴  王冰  姚云军  张晓通  江波 《遥感学报》2020,24(12):1433-1449

叶面积指数LAI（Leaf Area Index）是表征叶片疏密程度和冠层结构特征的重要植被参数，在气候变化、作物生长模型以及碳、水循环研究中发挥着重要作用。遥感是获取区域及全球尺度LAI的一个重要手段，当前LAI产品主要基于遥感数据反演得到，但是多数LAI产品算法并未考虑地形特征的影响，导致山地LAI遥感反演精度不确定性大。提高山地LAI遥感反演精度亟需考虑地形因子对冠层反射率的影响，其中山地冠层反射率模型和遥感数据地形校正是提升山地LAI遥感反演精度的关键。本文围绕山地LAI遥感反演理论与方法，综合分析了国内外山地冠层反射率模型和地形校正模型的研究进展，总结了目前山地LAI遥感反演存在的问题，并讨论了未来研究的发展趋势。  相似文献   

A new concept for estimating the influence of vegetation on throughfall kinetic energy using aerial laser scanning     

Johannes Antenor Senn  Fabian Ewald Fassnacht  Jana Eichel  Steffen Seitz  Sebastian Schmidtlein 《地球表面变化过程与地形》2020,45(7):1487-1498

Soil loss caused by erosion has enormous economic and social impacts. Splash effects of rainfall are an important driver of erosion processes; however, effects of vegetation on splash erosion are still not fully understood. Splash erosion processes under vegetation are investigated by means of throughfall kinetic energy (TKE). Previous studies on TKE utilized a heterogeneous set of plant and canopy parameters to assess vegetation's influence on erosion by rain splash but remained on individual plant- or plot-levels. In the present study we developed a method for the area-wide estimation of the influence of vegetation on TKE using remote sensing methods. In a literature review we identified key vegetation variables influencing splash erosion and developed a conceptual model to describe the interaction of vegetation and raindrops. Our model considers both amplifying and protecting effect of vegetation layers according to their height above the ground and aggregates them into a new indicator: the Vegetation Splash Factor (VSF). It is based on the proportional contribution of drips per layer, which can be calculated via the vegetation cover profile from airborne LiDAR datasets. In a case study, we calculated the VSF using a LiDAR dataset for La Campana National Park in central Chile. The studied catchment comprises a heterogeneous mosaic of vegetation layer combinations and types and is hence well suited to test the approach. We calculated a VSF map showing the relation between vegetation structure and its expected influence on TKE. Mean VSF was 1.42, indicating amplifying overall effect of vegetation on TKE that was present in 81% of the area. Values below 1 indicating a protective effect were calculated for 19% of the area. For future work, we recommend refining the weighting factor by calibration to local conditions using field-reference data and comparing the VSF with TKE field measurements. © 2020 The Authors. Earth Surface Processes and Landforms published by John Wiley & Sons Ltd  相似文献   

Monitoring soil surface roughness under growing winter wheat with low-altitude UAV sensing: Potential and limitations     

Nils Onnen  Anette Eltner  Goswin Heckrath  Kristof Van Oost 《地球表面变化过程与地形》2020,45(14):3747-3759

Soil surface roughness (SSR) is an important factor in controlling sediment and runoff generation, influencing directly a wide spectrum of erosion parameters. SSR is highly variable in time and space under natural conditions, and characterizing SSR to improve the parameterization of hydrological and erosion models has proved challenging. Our study uses recent technological and algorithmic developments in capturing and processing close aerial sensing data to evaluate how high-resolution imagery can assist the temporally and spatially explicit monitoring of SSR. We evaluated the evolution of SSR under natural rainfall and growing vegetation conditions on two arable fields in Denmark. Unmanned aerial vehicle (UAV) photogrammetry was used to monitor small field plots over 7 months after seeding of winter wheat following conventional and reduced tillage treatments. Field campaigns were conducted at least once a month from October until April, resulting in nine time steps of data acquisition. Structure from motion photogrammetry was used to derive high-resolution point clouds with an average ground sampling distance of 2.7 mm and a mean ground control point accuracy of 1.8 mm. A comprehensive workflow was developed to process the point clouds, including the detection of vegetation and the removal of vegetation-induced point cloud noise. Rasterized and filtered point clouds were then used to determine SSR geostatistically as the standard deviation of height, applying different kernel sizes and using semivariograms. The results showed an influence of kernel size on roughness, with a value range of 0.2–1 cm of average height deviation during the monitoring period. Semivariograms showed a measurable decrease in sill variance and an increase in range over time. This research demonstrated multiple challenges to measuring SSR with UAV under natural conditions with increasing vegetation cover. The proposed workflow represents a step forward in tackling those challenges and provides a knowledge base for future research. © 2020 John Wiley & Sons, Ltd.  相似文献   

国产高分辨率遥感影像农村公路核查方法及其应用     

蔡红玥  袁胜古  阳柯  王芳  盛光晓 《测绘通报》2020,(3):91-95

传统的农村公路核查需要人工实地抽查或通过GNSS设备进行信息采集验核，存在成本高、效率低等问题。遥感影像具有成像范围广、时效性高、成本低、能客观反映现实情况等优点。相比于传统方法，将遥感影像引入农村公路核查，能客观、准确、高效地对农村公路相关信息进行核查。本文基于国产高分辨率遥感影像，结合农村公路遥感核查业务，采用遥感影像道路提取算法，设计并实现了一种农村公路核查方法。将本方法应用于某中部省份农村公路遥感核查业务，实际应用表明该方法能有效提高现有农村公路遥感核查的工作效率。  相似文献   

一种基于GPU和内存映射文件的高分辨率遥感图像快速处理方法     

马秀丹  崔宾阁  钟勇  张永辉  费东 《海洋科学》2018,42(1):139-146

高分辨率遥感图像处理经常面临程序执行时间过长和内存空间不足的问题,虽然并行计算技术可以提高遥感图像的处理速度,但是无法降低算法占用的巨大内存空间。为了解决这一问题,本文提出了一种利用CUDA和内存映射文件的高分辨率遥感图像快速处理方法,并以K-Means算法为例进行了实现。其中,CUDA技术可以有效利用GPU强大的并行计算能力,而内存映射文件技术降低了磁盘I/O速度较慢对算法性能的影响。实验结果表明,本文方法比传统K-Means聚类算法计算速度提高了30倍左右,内存使用量降低了90%以上。  相似文献