Based on Lund and Shanklin’s work (1972), methods of calculating Probability of Cloud-Free Lines-of-Sight (PCFLOS), Persistence Probability of Cloud-Free Lines-of-Sight (PPCFLOS) and Recurrence Probability of Cloud-f...Based on Lund and Shanklin’s work (1972), methods of calculating Probability of Cloud-Free Lines-of-Sight (PCFLOS), Persistence Probability of Cloud-Free Lines-of-Sight (PPCFLOS) and Recurrence Probability of Cloud-free Lines-of-Sight (RPCFLOS) at given heights are presented. PCFLOS, PPCFLOS and RPCFLOS are calculated in Foshan, China by conventional observation data from 1961 to 1990. The conclusions are: (1) The higher the elevations, the smaller the PCFLOS and the larger the view angles, the larger the PCFLOS. (2) PPCFLOS and RPCFLOS decrease with the increase of elevation and the delay of time. (3) RPCFLOS is always equal to or larger than PPCFLOS at lag times.展开更多
The operational cloud-motion tracking technique fails to retrieve atmospheric motion vectors （AMVs） in areas lacking cloud; and while water vapor shown in water vapor imagery can be used, the heights assigned to the...The operational cloud-motion tracking technique fails to retrieve atmospheric motion vectors （AMVs） in areas lacking cloud; and while water vapor shown in water vapor imagery can be used, the heights assigned to the retrieved AMVs are mostly in the upper troposphere. As the noise-equivalent temperature difference （NEdT） performance of FY-2E split win- dow （10.3-11.5 μm, 11.6-12.8 μm） channels has been improved, the weak signals representing the spatial texture of water vapor and aerosols in cloud-free areas can be strengthened with algorithms based on the difference principle, and applied in calculating AMVs in the lower troposphere. This paper is a preliminary summary for this purpose, in which the principles and algorithm schemes for the temporal difference, split window difference and second-order difference （SD） methods are introduced. Results from simulation and cases experiments are reported in order to verify and evaluate the methods, based on comparison among retrievals and the ＂truth＂. The results show that all three algorithms, though not perfect in some cases, generally work well. Moreover, the SD method appears to be the best in suppressing the surface temperature influence and clarifying the spatial texture of water vapor and aerosols. The accuracy with respect to NCEP 800 hPa reanalysis data was found to be acceptable, as compared with the accuracy of the cloud motion vectors.展开更多
光学遥感影像不可避免会受到云的遮挡,导致影像应用性降低,遥感影像去云是近些年来的热门研究方向。插值算法利用一张或多张影像计算相似像素来重建缺失像素,但是最佳相似像素的确定和地物突变的高精度重建仍存在挑战。文章利用多时相...光学遥感影像不可避免会受到云的遮挡,导致影像应用性降低,遥感影像去云是近些年来的热门研究方向。插值算法利用一张或多张影像计算相似像素来重建缺失像素,但是最佳相似像素的确定和地物突变的高精度重建仍存在挑战。文章利用多时相影像时间相关性关系、空间关系和光谱关系,提出了基于时空谱约束的相似像素插值(SPISTS)的去云方法。该方法利用时域和谱域特征选择相似像素,并通过空间、时间相关性和光谱关系约束求得影像缺失值,最后通过正则化项对预测进行偏差改正。在3个不同地区进行实验并与加权线性回归(weighted linear regression,WLR)、时空加权回归(spatio-temporal weighted regression,STWR)、改进邻域相似像素插值(modified neighborhood similar pixel interpolation,MNSPI)结果进行比较。实验结果表明,该方法精度较高,能够减弱辐射差异带来的影响。展开更多
针对当前三维目标检测中存在的数据降采样难、特征提取不充分、感受野有限、候选包围盒回归质量不高等问题,基于3DSSD三维目标检测算法,提出了一种基于原始点云、单阶段、无锚框的三维目标检测算法RPV-SSD(random point voxel single st...针对当前三维目标检测中存在的数据降采样难、特征提取不充分、感受野有限、候选包围盒回归质量不高等问题,基于3DSSD三维目标检测算法,提出了一种基于原始点云、单阶段、无锚框的三维目标检测算法RPV-SSD(random point voxel single stage object detector),该算法由随机体素采样层、3D稀疏卷积层、特征聚合层、候选点生成层、区域建议网络层共五个部分组成,主要通过聚合随机体素采样的关键点逐点特征、体素稀疏卷积特征、鸟瞰图特征,进而实现对物体类别、3D包围盒以及物体朝向的预测。在KITTI数据集上的实验表明,该算法整体表现良好,不仅能够命中真值标签中的目标并且回归较好的包围盒,还能够从物体的不完整点云推测出物体的类别及其完整形状,提高目标检测性能。展开更多
文摘Based on Lund and Shanklin’s work (1972), methods of calculating Probability of Cloud-Free Lines-of-Sight (PCFLOS), Persistence Probability of Cloud-Free Lines-of-Sight (PPCFLOS) and Recurrence Probability of Cloud-free Lines-of-Sight (RPCFLOS) at given heights are presented. PCFLOS, PPCFLOS and RPCFLOS are calculated in Foshan, China by conventional observation data from 1961 to 1990. The conclusions are: (1) The higher the elevations, the smaller the PCFLOS and the larger the view angles, the larger the PCFLOS. (2) PPCFLOS and RPCFLOS decrease with the increase of elevation and the delay of time. (3) RPCFLOS is always equal to or larger than PPCFLOS at lag times.
基金supported by the National Natural Science Foundation of China(Grant Nos.41175035 and 40475018)the National Basic Research Program of China(Grant No.2009CB421502)a project funded by the Priority Academic Program Development of Jiangsu Higher Education Institutions
文摘The operational cloud-motion tracking technique fails to retrieve atmospheric motion vectors （AMVs） in areas lacking cloud; and while water vapor shown in water vapor imagery can be used, the heights assigned to the retrieved AMVs are mostly in the upper troposphere. As the noise-equivalent temperature difference （NEdT） performance of FY-2E split win- dow （10.3-11.5 μm, 11.6-12.8 μm） channels has been improved, the weak signals representing the spatial texture of water vapor and aerosols in cloud-free areas can be strengthened with algorithms based on the difference principle, and applied in calculating AMVs in the lower troposphere. This paper is a preliminary summary for this purpose, in which the principles and algorithm schemes for the temporal difference, split window difference and second-order difference （SD） methods are introduced. Results from simulation and cases experiments are reported in order to verify and evaluate the methods, based on comparison among retrievals and the ＂truth＂. The results show that all three algorithms, though not perfect in some cases, generally work well. Moreover, the SD method appears to be the best in suppressing the surface temperature influence and clarifying the spatial texture of water vapor and aerosols. The accuracy with respect to NCEP 800 hPa reanalysis data was found to be acceptable, as compared with the accuracy of the cloud motion vectors.
文摘光学遥感影像不可避免会受到云的遮挡,导致影像应用性降低,遥感影像去云是近些年来的热门研究方向。插值算法利用一张或多张影像计算相似像素来重建缺失像素,但是最佳相似像素的确定和地物突变的高精度重建仍存在挑战。文章利用多时相影像时间相关性关系、空间关系和光谱关系,提出了基于时空谱约束的相似像素插值(SPISTS)的去云方法。该方法利用时域和谱域特征选择相似像素,并通过空间、时间相关性和光谱关系约束求得影像缺失值,最后通过正则化项对预测进行偏差改正。在3个不同地区进行实验并与加权线性回归(weighted linear regression,WLR)、时空加权回归(spatio-temporal weighted regression,STWR)、改进邻域相似像素插值(modified neighborhood similar pixel interpolation,MNSPI)结果进行比较。实验结果表明,该方法精度较高,能够减弱辐射差异带来的影响。
文摘针对当前三维目标检测中存在的数据降采样难、特征提取不充分、感受野有限、候选包围盒回归质量不高等问题,基于3DSSD三维目标检测算法,提出了一种基于原始点云、单阶段、无锚框的三维目标检测算法RPV-SSD(random point voxel single stage object detector),该算法由随机体素采样层、3D稀疏卷积层、特征聚合层、候选点生成层、区域建议网络层共五个部分组成,主要通过聚合随机体素采样的关键点逐点特征、体素稀疏卷积特征、鸟瞰图特征,进而实现对物体类别、3D包围盒以及物体朝向的预测。在KITTI数据集上的实验表明,该算法整体表现良好,不仅能够命中真值标签中的目标并且回归较好的包围盒,还能够从物体的不完整点云推测出物体的类别及其完整形状,提高目标检测性能。
