Pub Date : 2019-07-01DOI: 10.1109/IGARSS.2019.8900595
G. Ferraioli, L. Denis, C. Deledalle, F. Tupin
In the last decades, several approaches for solving the Phase Unwrapping (PhU) problem using multi-channel Interferometric Synthetic Aperture Radar (InSAR) data have been developed. Many of the proposed approaches are based on statistical estimation theory, both classical and Bayesian. In particular, the statistical approaches based on the use of the whole complex multi-channel dataset have turned to be effective. The latter are based on the exploitation of the covariance matrix, which contains the parameters of interest. In this paper, the added value of the Non Local (NL) paradigm within the InSAR multi-channel PhU framework is investigated. The analysis of the impact of NL technique is performed using multi-channel realistic simulated data and X-band data.
{"title":"The Exploitation of the Non Local Paradigm for SAR 3d Reconstruction","authors":"G. Ferraioli, L. Denis, C. Deledalle, F. Tupin","doi":"10.1109/IGARSS.2019.8900595","DOIUrl":"https://doi.org/10.1109/IGARSS.2019.8900595","url":null,"abstract":"In the last decades, several approaches for solving the Phase Unwrapping (PhU) problem using multi-channel Interferometric Synthetic Aperture Radar (InSAR) data have been developed. Many of the proposed approaches are based on statistical estimation theory, both classical and Bayesian. In particular, the statistical approaches based on the use of the whole complex multi-channel dataset have turned to be effective. The latter are based on the exploitation of the covariance matrix, which contains the parameters of interest. In this paper, the added value of the Non Local (NL) paradigm within the InSAR multi-channel PhU framework is investigated. The analysis of the impact of NL technique is performed using multi-channel realistic simulated data and X-band data.","PeriodicalId":13262,"journal":{"name":"IGARSS 2019 - 2019 IEEE International Geoscience and Remote Sensing Symposium","volume":"262 1","pages":"5193-5196"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75850107","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-07-01DOI: 10.1109/IGARSS.2019.8898849
Xiaoyong Bian, Chunfang Chen, Chunhua Deng, Ruiyao Liu, Q. Du
High-resolution scene classification is a fundamental yet challenging problem due to rich image variations in viewpoint, object pose and spatial resolution, etc, which results in large within-class diversity and high between-class similarity. In the paper we focus on tackling the problem of how to learn appropriate feature representation for high-resolution scene classification. To achieve better scene representation, we proposed a combined CNN feature learning framework in multi-scale multi-layer based Gaussian coding (mSmL-Gcoding) manner. In addition, a novel feature coding with Gaussian descriptor is introduced to enhance the discriminative ability of CNN features. Experimental results on two publicly available challenging scene datasets validated that the effectiveness of our method and found it compared favorably with state-of-the-arts.
{"title":"Hierarchical Deep Feature Representation for High-Resolution Scene Classification","authors":"Xiaoyong Bian, Chunfang Chen, Chunhua Deng, Ruiyao Liu, Q. Du","doi":"10.1109/IGARSS.2019.8898849","DOIUrl":"https://doi.org/10.1109/IGARSS.2019.8898849","url":null,"abstract":"High-resolution scene classification is a fundamental yet challenging problem due to rich image variations in viewpoint, object pose and spatial resolution, etc, which results in large within-class diversity and high between-class similarity. In the paper we focus on tackling the problem of how to learn appropriate feature representation for high-resolution scene classification. To achieve better scene representation, we proposed a combined CNN feature learning framework in multi-scale multi-layer based Gaussian coding (mSmL-Gcoding) manner. In addition, a novel feature coding with Gaussian descriptor is introduced to enhance the discriminative ability of CNN features. Experimental results on two publicly available challenging scene datasets validated that the effectiveness of our method and found it compared favorably with state-of-the-arts.","PeriodicalId":13262,"journal":{"name":"IGARSS 2019 - 2019 IEEE International Geoscience and Remote Sensing Symposium","volume":"3 1","pages":"517-520"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75863380","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-07-01DOI: 10.1109/IGARSS.2019.8898484
Menghao Ji, B. Tang
The Fraunhofer line discrimination (FLD) principle is widely used for retrieving solar-induced chlorophyll fluorescence (SIF), which assumes that the spectral reflectance is smooth and can be modeled using simply mathematical function. However, the changes in the sun and observation geometry and atmospheric properties result in the ‘hump’ or ‘dip’ of the reflectance spectrum in the oxygen A-band. This leads to overestimations or underestimations in the SIF retrieval. The principal component analysis (PCA) algorithm is one of the main approaches used for satellite-based SIF retrieval, which can acquire reflectance characteristic information due to directional effect with large datasets. This paper attempts to test whether the errors caused by FLD method can be eliminated using the PCA algorithm. The results show that the PCA algorithm performs well in all conditions, with root mean square error less than 0.005, indicating that the bias caused by the changes in sun and observation geometry could be eliminated with PCA algorithm.
{"title":"Retreval of Solar-Induced Chlorohyll Fluoresence with Principal Component Ananlysis Method","authors":"Menghao Ji, B. Tang","doi":"10.1109/IGARSS.2019.8898484","DOIUrl":"https://doi.org/10.1109/IGARSS.2019.8898484","url":null,"abstract":"The Fraunhofer line discrimination (FLD) principle is widely used for retrieving solar-induced chlorophyll fluorescence (SIF), which assumes that the spectral reflectance is smooth and can be modeled using simply mathematical function. However, the changes in the sun and observation geometry and atmospheric properties result in the ‘hump’ or ‘dip’ of the reflectance spectrum in the oxygen A-band. This leads to overestimations or underestimations in the SIF retrieval. The principal component analysis (PCA) algorithm is one of the main approaches used for satellite-based SIF retrieval, which can acquire reflectance characteristic information due to directional effect with large datasets. This paper attempts to test whether the errors caused by FLD method can be eliminated using the PCA algorithm. The results show that the PCA algorithm performs well in all conditions, with root mean square error less than 0.005, indicating that the bias caused by the changes in sun and observation geometry could be eliminated with PCA algorithm.","PeriodicalId":13262,"journal":{"name":"IGARSS 2019 - 2019 IEEE International Geoscience and Remote Sensing Symposium","volume":"27 1","pages":"1955-1958"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74450684","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-07-01DOI: 10.1109/IGARSS.2019.8898180
Jiaxin Tang, Fan Zhang, Yongsheng Zhou, Q. Yin, Wei Hu
In this paper, we improve the Siamese Networks for SAR target few-shot learning. SAR target recognition is an important branch of SAR application. It can efficiently extract target category information from complex SAR images and help humans quickly understand SAR images. However, many successful machine learning methods require large amounts of annotated data. So, few-shot learning is always a topical challenge for machine learning. We apply Siamese Networks to SAR target recognition with limited data and improved it. Our model consists of CNN encoder, similarity discriminator and classifier. Relevantly, it has two inputs and three outputs. CNN encoder is constrained by similarity discriminator and classifier. Furthermore, the larger difference from the Siamese Network is that the target category is outputted by the classifier, not by the similarity discriminator. Our method not only makes use of the advantage of metric learning to improve the accuracy of SAR target recognition with limited data, but also significantly reduces the prediction time consumption for the model based on metric learning. In the ten categories military vehicle classification task, there are only five samples for each category and a total of 2425 testing samples. Our method outperforms A-ConvNet and Siamese Networks by 15.8% and 8.41%. The prediction time consumption of Siamese Networks is 114.832s, while that of our method is 1.172s.
{"title":"A Fast Inference Networks for SAR Target Few-Shot Learning Based on Improved Siamese Networks","authors":"Jiaxin Tang, Fan Zhang, Yongsheng Zhou, Q. Yin, Wei Hu","doi":"10.1109/IGARSS.2019.8898180","DOIUrl":"https://doi.org/10.1109/IGARSS.2019.8898180","url":null,"abstract":"In this paper, we improve the Siamese Networks for SAR target few-shot learning. SAR target recognition is an important branch of SAR application. It can efficiently extract target category information from complex SAR images and help humans quickly understand SAR images. However, many successful machine learning methods require large amounts of annotated data. So, few-shot learning is always a topical challenge for machine learning. We apply Siamese Networks to SAR target recognition with limited data and improved it. Our model consists of CNN encoder, similarity discriminator and classifier. Relevantly, it has two inputs and three outputs. CNN encoder is constrained by similarity discriminator and classifier. Furthermore, the larger difference from the Siamese Network is that the target category is outputted by the classifier, not by the similarity discriminator. Our method not only makes use of the advantage of metric learning to improve the accuracy of SAR target recognition with limited data, but also significantly reduces the prediction time consumption for the model based on metric learning. In the ten categories military vehicle classification task, there are only five samples for each category and a total of 2425 testing samples. Our method outperforms A-ConvNet and Siamese Networks by 15.8% and 8.41%. The prediction time consumption of Siamese Networks is 114.832s, while that of our method is 1.172s.","PeriodicalId":13262,"journal":{"name":"IGARSS 2019 - 2019 IEEE International Geoscience and Remote Sensing Symposium","volume":"16 1","pages":"1212-1215"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74499537","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-07-01DOI: 10.1109/IGARSS.2019.8899064
Jingyi Jiang, M. Weiss, Shouyang Liu, F. Baret
Green Area Index (GAI) and Leaf Chlorophyll Content (LCC) are key variables that reflect the potential growth of the canopy. In the past decades, the retrieval of these variables from remote sensing data to generate operational products at high spatial resolution (lower than decametric) was mainly based on 1D radiative transfer model inversion. However, due to the recent advances in computational facility, it is now possible to invert 3D radiative transfer models to improve the operational product accuracy. The use of 3D models allows taking into account more realistic canopy architectures than when using the turbid medium assumption from the 1D radiative transfer models. In this study, we demonstrate the gain in accuracy when inverting crop specific using 3D radiative transfer models as compared to a generic algorithm based on 1D model. We investigate two crops characterized by contrasted architectures along the vegetation cycle, e.g. wheat and maize.
{"title":"The impact of canopy structure assumption on the retrieval of GAI and Leaf Chlorophyll Content for wheat and maize crops","authors":"Jingyi Jiang, M. Weiss, Shouyang Liu, F. Baret","doi":"10.1109/IGARSS.2019.8899064","DOIUrl":"https://doi.org/10.1109/IGARSS.2019.8899064","url":null,"abstract":"Green Area Index (GAI) and Leaf Chlorophyll Content (LCC) are key variables that reflect the potential growth of the canopy. In the past decades, the retrieval of these variables from remote sensing data to generate operational products at high spatial resolution (lower than decametric) was mainly based on 1D radiative transfer model inversion. However, due to the recent advances in computational facility, it is now possible to invert 3D radiative transfer models to improve the operational product accuracy. The use of 3D models allows taking into account more realistic canopy architectures than when using the turbid medium assumption from the 1D radiative transfer models. In this study, we demonstrate the gain in accuracy when inverting crop specific using 3D radiative transfer models as compared to a generic algorithm based on 1D model. We investigate two crops characterized by contrasted architectures along the vegetation cycle, e.g. wheat and maize.","PeriodicalId":13262,"journal":{"name":"IGARSS 2019 - 2019 IEEE International Geoscience and Remote Sensing Symposium","volume":"18 1","pages":"7216-7219"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74509933","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Finding a suitable parking position often leads to much traffic pressure and time consumption in a busy parking lot. An auxiliary parking method based on automotive millimeter wave SAR is proposed in this paper. Firstly, Maximally Stable Extremal Region (MSER) method is utilized to extract the candidate regions occupied by parked vehicles from the millimeter wave SAR images. Then, in order to eliminate the false alarm candidate regions, we employ the morphological filter and utilize the centroid position to further refine the candidate regions. Thirdly, the difference in width-to-height ratio of the candidate regions is exploited to distinguish the parking directions of the cars. After that, the available parking spaces are located according to the parking direction. Finally, further remove the spaces occupied by obstacles, and plan reasonable parking routes. Experimental results based on measured data show that the proposed method has outstanding detection and parking route planning performance in different scenes.
{"title":"An Auxiliary Parking Method Based on Automotive Millimeter wave SAR","authors":"Rufei Wang, Jifang Pei, Yongchao Zhang, Minghui Li, Yulin Huang, Junjie Wu","doi":"10.1109/IGARSS.2019.8898521","DOIUrl":"https://doi.org/10.1109/IGARSS.2019.8898521","url":null,"abstract":"Finding a suitable parking position often leads to much traffic pressure and time consumption in a busy parking lot. An auxiliary parking method based on automotive millimeter wave SAR is proposed in this paper. Firstly, Maximally Stable Extremal Region (MSER) method is utilized to extract the candidate regions occupied by parked vehicles from the millimeter wave SAR images. Then, in order to eliminate the false alarm candidate regions, we employ the morphological filter and utilize the centroid position to further refine the candidate regions. Thirdly, the difference in width-to-height ratio of the candidate regions is exploited to distinguish the parking directions of the cars. After that, the available parking spaces are located according to the parking direction. Finally, further remove the spaces occupied by obstacles, and plan reasonable parking routes. Experimental results based on measured data show that the proposed method has outstanding detection and parking route planning performance in different scenes.","PeriodicalId":13262,"journal":{"name":"IGARSS 2019 - 2019 IEEE International Geoscience and Remote Sensing Symposium","volume":"224 1","pages":"2503-2506"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74671832","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-07-01DOI: 10.1109/IGARSS.2019.8899101
Jun Su Kim, K. Papathanassiou
A methodology that allows the separation of Faraday Rotation (FR) distortion from system induced distortion for the calibration of spaceborne polarimetric data is proposed and discussed. The separation is based on the azimuth variation of the FR and relies on the assumption that the antenna patterns are sufficiently well characterized.
{"title":"Polarimetric Calibration of Spaceborne SAR Data in The Presence of the Ionosphere by Means of Azimuth Sub-Bands","authors":"Jun Su Kim, K. Papathanassiou","doi":"10.1109/IGARSS.2019.8899101","DOIUrl":"https://doi.org/10.1109/IGARSS.2019.8899101","url":null,"abstract":"A methodology that allows the separation of Faraday Rotation (FR) distortion from system induced distortion for the calibration of spaceborne polarimetric data is proposed and discussed. The separation is based on the azimuth variation of the FR and relies on the assumption that the antenna patterns are sufficiently well characterized.","PeriodicalId":13262,"journal":{"name":"IGARSS 2019 - 2019 IEEE International Geoscience and Remote Sensing Symposium","volume":"90 1","pages":"5278-5280"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74759537","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-07-01DOI: 10.1109/IGARSS.2019.8898733
Xiaoyu Yan, Jie Chen, H. Nies, H. Zeng, O. Loffeld
Real-time imaging products using Spaceborne Trinodal Pendulum synthetic aperture radar formation can provide valuable information in certain applications. The onboard orbit determination data of the spaceborne SAR platform is essential for the SAR imaging procedure. For real-time SAR imaging, the onboard orbit determination data is relatively low in accuracy compared with the orbit data obtained by off-line processing for the focusing of SAR data. The influence of errors in onboard real-time orbit determination data on SAR image quality should be considered. This paper proposes a Monte Carlo simulation model for inspecting the influence of onboard orbit determination data on imaging quality. This simulation model and its result may be helpful for the development of SAR real-time imaging focusing on providing terrain change information in a short time.
{"title":"Analysis of Quadratic Phase Error Introduced by Orbit Determination in Spaceborne Trinodal Pendulum Sar Formation Real-Time Imaging with Monte Carlo Simulation","authors":"Xiaoyu Yan, Jie Chen, H. Nies, H. Zeng, O. Loffeld","doi":"10.1109/IGARSS.2019.8898733","DOIUrl":"https://doi.org/10.1109/IGARSS.2019.8898733","url":null,"abstract":"Real-time imaging products using Spaceborne Trinodal Pendulum synthetic aperture radar formation can provide valuable information in certain applications. The onboard orbit determination data of the spaceborne SAR platform is essential for the SAR imaging procedure. For real-time SAR imaging, the onboard orbit determination data is relatively low in accuracy compared with the orbit data obtained by off-line processing for the focusing of SAR data. The influence of errors in onboard real-time orbit determination data on SAR image quality should be considered. This paper proposes a Monte Carlo simulation model for inspecting the influence of onboard orbit determination data on imaging quality. This simulation model and its result may be helpful for the development of SAR real-time imaging focusing on providing terrain change information in a short time.","PeriodicalId":13262,"journal":{"name":"IGARSS 2019 - 2019 IEEE International Geoscience and Remote Sensing Symposium","volume":"106 1","pages":"8582-8585"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74880388","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-07-01DOI: 10.1109/IGARSS.2019.8899035
Xin-Yi Tong, Jihao Yin, Limin Wu, Hui Qv
A global self-labeled distribution analysis (GSLDA) for hyperspectral image (HSI) band selection is proposed in this paper, which focuses on an unsupervised method to ascertain the band discrimination. In order to generate the band labels for further analysis, the concept of the local minimum spanning forest (LMSF) is introduced into the construction of the global self-labeled band partitions based on graph theory. Meanwhile, the novel scoring strategy of triple-density indexes is applied to analyze the labeled-band distribution for determining the selected band subset with clear discrimination. The feasibility of the proposed method is evaluated on real hyperspectral data and the experiment results show a competitive good performance, which demonstrates that the selected bands hold apparent global discrimination and robust noise immunity.
{"title":"Global Self-Labeled Distribution Analysis for Hyperspectral Band Selection","authors":"Xin-Yi Tong, Jihao Yin, Limin Wu, Hui Qv","doi":"10.1109/IGARSS.2019.8899035","DOIUrl":"https://doi.org/10.1109/IGARSS.2019.8899035","url":null,"abstract":"A global self-labeled distribution analysis (GSLDA) for hyperspectral image (HSI) band selection is proposed in this paper, which focuses on an unsupervised method to ascertain the band discrimination. In order to generate the band labels for further analysis, the concept of the local minimum spanning forest (LMSF) is introduced into the construction of the global self-labeled band partitions based on graph theory. Meanwhile, the novel scoring strategy of triple-density indexes is applied to analyze the labeled-band distribution for determining the selected band subset with clear discrimination. The feasibility of the proposed method is evaluated on real hyperspectral data and the experiment results show a competitive good performance, which demonstrates that the selected bands hold apparent global discrimination and robust noise immunity.","PeriodicalId":13262,"journal":{"name":"IGARSS 2019 - 2019 IEEE International Geoscience and Remote Sensing Symposium","volume":"8 1","pages":"3792-3795"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74927717","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2019-07-01DOI: 10.1109/IGARSS.2019.8899896
Jianjun Liu, Hao Chen, Songze Tang, Jinlong Yang, Hong Yan
In this paper, we investigate the ridge regression for multivariate labels by modelling each pixel and its surrounding pixels as a 3D tensor, and thereby propose a tensor ridge regression approach (TRR) for spatial-spectral hyperspectral image classification. Compared with the traditional ridge regression model, not only the spatial information is incorporated, but also the intrinsic spatial-spectral structure is captured. Moreover, the proposed TRR method is universal that it can be adopted to deal with the fusion of multiscale features for classification purpose. Experiment results conducted on two hyperspectral scenes demonstrate the effectiveness of the proposed method.
{"title":"Hyperspectral Image Classification Via Tensor Ridge Regression","authors":"Jianjun Liu, Hao Chen, Songze Tang, Jinlong Yang, Hong Yan","doi":"10.1109/IGARSS.2019.8899896","DOIUrl":"https://doi.org/10.1109/IGARSS.2019.8899896","url":null,"abstract":"In this paper, we investigate the ridge regression for multivariate labels by modelling each pixel and its surrounding pixels as a 3D tensor, and thereby propose a tensor ridge regression approach (TRR) for spatial-spectral hyperspectral image classification. Compared with the traditional ridge regression model, not only the spatial information is incorporated, but also the intrinsic spatial-spectral structure is captured. Moreover, the proposed TRR method is universal that it can be adopted to deal with the fusion of multiscale features for classification purpose. Experiment results conducted on two hyperspectral scenes demonstrate the effectiveness of the proposed method.","PeriodicalId":13262,"journal":{"name":"IGARSS 2019 - 2019 IEEE International Geoscience and Remote Sensing Symposium","volume":"307 2 1","pages":"1156-1159"},"PeriodicalIF":0.0,"publicationDate":"2019-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"72946768","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：