The automatic registration of multi-source remote sensing images (RSI) is a research hotspot of remote sensing image preprocessing currently. A special automatic image registration module named the Image Autosync has ...The automatic registration of multi-source remote sensing images (RSI) is a research hotspot of remote sensing image preprocessing currently. A special automatic image registration module named the Image Autosync has been embedded into the ERDAS IMAGINE software of version 9.0 and above. The registration accuracies of the module verified for the remote sensing images obtained from different platforms or their different spatial resolution. Four tested registration experiments are discussed in this article to analyze the accuracy differences based on the remote sensing data which have different spatial resolution. The impact factors inducing the differences of registration accuracy are also analyzed.展开更多
Image registration is an indispensable component in multi-source remote sensing image processing. In this paper, we put forward a remote sensing image registration method by including an improved multi-scale and multi...Image registration is an indispensable component in multi-source remote sensing image processing. In this paper, we put forward a remote sensing image registration method by including an improved multi-scale and multi-direction Harris algorithm and a novel compound feature. Multi-scale circle Gaussian combined invariant moments and multi-direction gray level co-occurrence matrix are extracted as features for image matching. The proposed algorithm is evaluated on numerous multi-source remote sensor images with noise and illumination changes. Extensive experimental studies prove that our proposed method is capable of receiving stable and even distribution of key points as well as obtaining robust and accurate correspondence matches. It is a promising scheme in multi-source remote sensing image registration.展开更多
Snow depth （SD） is a key parameter for research into global climate changes and land surface processes. A method was developed to obtain daily SD images at a higher 4 km spatial resolution and higher precision with ...Snow depth （SD） is a key parameter for research into global climate changes and land surface processes. A method was developed to obtain daily SD images at a higher 4 km spatial resolution and higher precision with SD measurements from in situ observations and passive microwave remote sensing of Advanced Microwave Scanning Radiometer-EOS （AMSR-E） and snow cover measurements of the Interactive Multisensor Snow and Ice Mapping System （IMS）. AMSR-E SD at 25 km spatial resolution was retrieved from AMSR-E products of snow density and snow water equivalent and then corrected using the SD from in situ observations and IMS snow cover. Corrected AMSR-E SD images were then resampled to act as ＂virtual＂ in situ observations to combine with the real in situ observations to interpolate at 4 km spatial resolution SD using the Cressman method. Finally, daily SD data generation for several regions of China demonstrated that the method is well suited to the generation of higher spatial resolution SD data in regions with a lower Digital Elevation Model （DEM） but not so well suited to regions at high altitude and with an undulating terrain, such as the Tibetan Plateau. Analysis of the longer time period SD data generation for January between 2003 and 2010 in northern Xinjiang also demonstrated the feasibility of the method.展开更多
多模态语义分割网络可以利用不同模态中的互补数据提升分割精度,但现有的多模语义分割模型多为将两种模态数据简单拼接起来,忽略了不同模态数据在高低频中的特征特性,导致跨模态特征提取不充分,融合不理想。针对上述问题,提出了将IRRG...多模态语义分割网络可以利用不同模态中的互补数据提升分割精度,但现有的多模语义分割模型多为将两种模态数据简单拼接起来,忽略了不同模态数据在高低频中的特征特性,导致跨模态特征提取不充分,融合不理想。针对上述问题,提出了将IRRG图像与DSM图像融合的遥感图像语义分割网络LHFNet(low feature and high feature fusion network)。针对各模态图像低频相关的结构特征,设计了低级特征提取加强模块,以加强不同模态特征的提取;基于各模态图像高频相互独立的细节特征,设计了高级特征融合模块,以指导不同模态的特征融合;针对高低频图像特征之间的语义鸿沟,设计了全局空洞空间金字塔池化模块跳跃连接高低频信息,以增强高低频图像特征之间的信息交互。通过基于ISPRS提供的Vaihingen和Potsdam数据集上的实验表明,LHFNet分别取得了88.17%和90.53%的全局精确度,相较于SegNet、DeepLabv3+等单模分割网络和RedNet、TSNet等多模RGB-D分割网络都具有更高的分割精确度。展开更多
文摘The automatic registration of multi-source remote sensing images (RSI) is a research hotspot of remote sensing image preprocessing currently. A special automatic image registration module named the Image Autosync has been embedded into the ERDAS IMAGINE software of version 9.0 and above. The registration accuracies of the module verified for the remote sensing images obtained from different platforms or their different spatial resolution. Four tested registration experiments are discussed in this article to analyze the accuracy differences based on the remote sensing data which have different spatial resolution. The impact factors inducing the differences of registration accuracy are also analyzed.
基金supported by National Nature Science Foundation of China (Nos. 61462046 and 61762052)Natural Science Foundation of Jiangxi Province (Nos. 20161BAB202049 and 20161BAB204172)+2 种基金the Bidding Project of the Key Laboratory of Watershed Ecology and Geographical Environment Monitoring, NASG (Nos. WE2016003, WE2016013 and WE2016015)the Science and Technology Research Projects of Jiangxi Province Education Department (Nos. GJJ160741, GJJ170632 and GJJ170633)the Art Planning Project of Jiangxi Province (Nos. YG2016250 and YG2017381)
文摘Image registration is an indispensable component in multi-source remote sensing image processing. In this paper, we put forward a remote sensing image registration method by including an improved multi-scale and multi-direction Harris algorithm and a novel compound feature. Multi-scale circle Gaussian combined invariant moments and multi-direction gray level co-occurrence matrix are extracted as features for image matching. The proposed algorithm is evaluated on numerous multi-source remote sensor images with noise and illumination changes. Extensive experimental studies prove that our proposed method is capable of receiving stable and even distribution of key points as well as obtaining robust and accurate correspondence matches. It is a promising scheme in multi-source remote sensing image registration.
基金Meteorological Research in the Public Interest,No.GYHY201106014Beijing Nova Program,No.2010B037China Special Fund for the National High Technology Research and Development Program of China(863 Program),No.412230
文摘Snow depth （SD） is a key parameter for research into global climate changes and land surface processes. A method was developed to obtain daily SD images at a higher 4 km spatial resolution and higher precision with SD measurements from in situ observations and passive microwave remote sensing of Advanced Microwave Scanning Radiometer-EOS （AMSR-E） and snow cover measurements of the Interactive Multisensor Snow and Ice Mapping System （IMS）. AMSR-E SD at 25 km spatial resolution was retrieved from AMSR-E products of snow density and snow water equivalent and then corrected using the SD from in situ observations and IMS snow cover. Corrected AMSR-E SD images were then resampled to act as ＂virtual＂ in situ observations to combine with the real in situ observations to interpolate at 4 km spatial resolution SD using the Cressman method. Finally, daily SD data generation for several regions of China demonstrated that the method is well suited to the generation of higher spatial resolution SD data in regions with a lower Digital Elevation Model （DEM） but not so well suited to regions at high altitude and with an undulating terrain, such as the Tibetan Plateau. Analysis of the longer time period SD data generation for January between 2003 and 2010 in northern Xinjiang also demonstrated the feasibility of the method.
文摘多模态语义分割网络可以利用不同模态中的互补数据提升分割精度,但现有的多模语义分割模型多为将两种模态数据简单拼接起来,忽略了不同模态数据在高低频中的特征特性,导致跨模态特征提取不充分,融合不理想。针对上述问题,提出了将IRRG图像与DSM图像融合的遥感图像语义分割网络LHFNet(low feature and high feature fusion network)。针对各模态图像低频相关的结构特征,设计了低级特征提取加强模块,以加强不同模态特征的提取;基于各模态图像高频相互独立的细节特征,设计了高级特征融合模块,以指导不同模态的特征融合;针对高低频图像特征之间的语义鸿沟,设计了全局空洞空间金字塔池化模块跳跃连接高低频信息,以增强高低频图像特征之间的信息交互。通过基于ISPRS提供的Vaihingen和Potsdam数据集上的实验表明,LHFNet分别取得了88.17%和90.53%的全局精确度,相较于SegNet、DeepLabv3+等单模分割网络和RedNet、TSNet等多模RGB-D分割网络都具有更高的分割精确度。
