Pub Date : 2018-01-01DOI: 10.1109/LGRS.2017.2777962
Zhihuo Xu, Quan Shi
To improve traffic safety, millimeter wave radars have been widely used for sensing traffic environment. As radars also operate on a narrow small road and in the same frequency band, mutual interference between different automotive radars that arises cannot be easily reduced by frequency or polarization diversity. This letter presents novel orthogonal noise waveforms to reduce such neighboring interferences. First, the spectral density distribution function of the proposed waveforms is defined by using an optimized Kaiser function. Subsequently, the phases of the noise waveforms are formulated as a problem of phase retrieval and are explored. Thanks to nonuniqueness solutions, the proposed method generates the orthogonal signals with a good random phase diversity. The proposed method was tested on a representative scenario for interference reduction. The experimental results show that the proposed method can produce visually convincing radar images, and the signal-to-interference and noise ratio is better than the existing methods.
{"title":"Interference Mitigation for Automotive Radar Using Orthogonal Noise Waveforms","authors":"Zhihuo Xu, Quan Shi","doi":"10.1109/LGRS.2017.2777962","DOIUrl":"https://doi.org/10.1109/LGRS.2017.2777962","url":null,"abstract":"To improve traffic safety, millimeter wave radars have been widely used for sensing traffic environment. As radars also operate on a narrow small road and in the same frequency band, mutual interference between different automotive radars that arises cannot be easily reduced by frequency or polarization diversity. This letter presents novel orthogonal noise waveforms to reduce such neighboring interferences. First, the spectral density distribution function of the proposed waveforms is defined by using an optimized Kaiser function. Subsequently, the phases of the noise waveforms are formulated as a problem of phase retrieval and are explored. Thanks to nonuniqueness solutions, the proposed method generates the orthogonal signals with a good random phase diversity. The proposed method was tested on a representative scenario for interference reduction. The experimental results show that the proposed method can produce visually convincing radar images, and the signal-to-interference and noise ratio is better than the existing methods.","PeriodicalId":13046,"journal":{"name":"IEEE Geoscience and Remote Sensing Letters","volume":"15 1","pages":"137-141"},"PeriodicalIF":4.8,"publicationDate":"2018-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1109/LGRS.2017.2777962","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"62472090","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-01-01DOI: 10.1109/LGRS.2017.2777450
Jie Geng, Xiaorui Ma, Jianchao Fan, Hongyu Wang
The classification of polarimetric synthetic aperture radar (PolSAR) image is of crucial significance for SAR applications. In this letter, a superpixel restrained deep neural network with multiple decisions (SRDNN-MDs) is proposed for PolSAR image classification, which not only extracts effective superpixel spatial features and degrades the influence of speckle noises but also deals with the limited training samples. First, the polarimetric features of coherency matrix and Yamaguchi decomposition are extracted as initial features, and superpixel segmentation is conducted on the Pauli color-coded image to acquire the superpixel averaged features. Then, an SRDNN based on sparse autoencoders is proposed to capture superpixel correlative features and reduce speckle noises. After that, MDs, including nonlocal decision and local decision, are developed to select credible testing samples. Finally, our deep network is updated by the extended training set to yield the final classification map. Experimental results demonstrate that the proposed SRDNN-MD yields higher accuracies compared with other related approaches, which indicate that the proposed method is able to capture superpixel correlative information and adds the information of unlabeled samples to improve the classification performance.
{"title":"Semisupervised Classification of Polarimetric SAR Image via Superpixel Restrained Deep Neural Network","authors":"Jie Geng, Xiaorui Ma, Jianchao Fan, Hongyu Wang","doi":"10.1109/LGRS.2017.2777450","DOIUrl":"https://doi.org/10.1109/LGRS.2017.2777450","url":null,"abstract":"The classification of polarimetric synthetic aperture radar (PolSAR) image is of crucial significance for SAR applications. In this letter, a superpixel restrained deep neural network with multiple decisions (SRDNN-MDs) is proposed for PolSAR image classification, which not only extracts effective superpixel spatial features and degrades the influence of speckle noises but also deals with the limited training samples. First, the polarimetric features of coherency matrix and Yamaguchi decomposition are extracted as initial features, and superpixel segmentation is conducted on the Pauli color-coded image to acquire the superpixel averaged features. Then, an SRDNN based on sparse autoencoders is proposed to capture superpixel correlative features and reduce speckle noises. After that, MDs, including nonlocal decision and local decision, are developed to select credible testing samples. Finally, our deep network is updated by the extended training set to yield the final classification map. Experimental results demonstrate that the proposed SRDNN-MD yields higher accuracies compared with other related approaches, which indicate that the proposed method is able to capture superpixel correlative information and adds the information of unlabeled samples to improve the classification performance.","PeriodicalId":13046,"journal":{"name":"IEEE Geoscience and Remote Sensing Letters","volume":"4 1","pages":"122-126"},"PeriodicalIF":4.8,"publicationDate":"2018-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1109/LGRS.2017.2777450","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"62471947","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-01-01DOI: 10.1109/LGRS.2017.2778045
Naihao Liu, Jing-Hua Gao, Bo Zhang, Fangyu Li, Qian Wang
The S transform (ST) is one of the most commonly used time–frequency (TF) analysis algorithms and is commonly used in assisting reservoir characterization and hydrocarbon detection. Unfortunately, the TF spectrum obtained by the ST has a low temporal resolution at low frequencies, which lowers its ability in thin beds and channels detection. In this letter, we propose a three parameters ST (TPST) to optimize the TF resolution flexibly. To demonstrate the validity and effectiveness of the TPST, we first apply it to a synthetic data and a synthetic seismic trace and then to a filed data. Synthetic data examples show that this TPST achieves an optimized TF resolution, compared with the standard ST and modified ST with two parameters. Field data experiments illustrate that the TPST is superior to the ST in highlighting the channel edges. The lateral continuity of the frequency slice produced by the TPST is more continuous than that of the ST.
{"title":"Time–Frequency Analysis of Seismic Data Using a Three Parameters S Transform","authors":"Naihao Liu, Jing-Hua Gao, Bo Zhang, Fangyu Li, Qian Wang","doi":"10.1109/LGRS.2017.2778045","DOIUrl":"https://doi.org/10.1109/LGRS.2017.2778045","url":null,"abstract":"The S transform (ST) is one of the most commonly used time–frequency (TF) analysis algorithms and is commonly used in assisting reservoir characterization and hydrocarbon detection. Unfortunately, the TF spectrum obtained by the ST has a low temporal resolution at low frequencies, which lowers its ability in thin beds and channels detection. In this letter, we propose a three parameters ST (TPST) to optimize the TF resolution flexibly. To demonstrate the validity and effectiveness of the TPST, we first apply it to a synthetic data and a synthetic seismic trace and then to a filed data. Synthetic data examples show that this TPST achieves an optimized TF resolution, compared with the standard ST and modified ST with two parameters. Field data experiments illustrate that the TPST is superior to the ST in highlighting the channel edges. The lateral continuity of the frequency slice produced by the TPST is more continuous than that of the ST.","PeriodicalId":13046,"journal":{"name":"IEEE Geoscience and Remote Sensing Letters","volume":"15 1","pages":"142-146"},"PeriodicalIF":4.8,"publicationDate":"2018-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1109/LGRS.2017.2778045","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"62472121","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-01-01DOI: 10.1109/LGRS.2017.2777598
Guowei Zhang, Jinghuai Gao
Attenuation is a fundamental mechanism as seismic wave propagates through the earth. The loss of high-frequency energy and concomitant phase distortion can be compensated by inverse ${Q}$ filtering to enhance the resolution of seismic data. Since the attenuation process depends on time and frequency, it is routinely performed in the time–frequency domain. The synchrosqueezing transform (SST), which provides highly localized time–frequency representations for the nonstationary signals due to reduced spectral smearing, is applied to implement the inverse ${Q}$ filtering scheme. However, the amplitude compensation process is unstable because energy amplification is involved. To stabilize it, the amplitude compensation is regarded as an inverse problem with an L1-norm regularization term in the SST domain. The iteratively reweighted least-squares algorithm is used to solve the regularized inverse problem. Synthetic and real data examples illustrate the stability and effectiveness of the proposed method.
{"title":"Inversion-Driven Attenuation Compensation Using Synchrosqueezing Transform","authors":"Guowei Zhang, Jinghuai Gao","doi":"10.1109/LGRS.2017.2777598","DOIUrl":"https://doi.org/10.1109/LGRS.2017.2777598","url":null,"abstract":"Attenuation is a fundamental mechanism as seismic wave propagates through the earth. The loss of high-frequency energy and concomitant phase distortion can be compensated by inverse <inline-formula> <tex-math notation=\"LaTeX\">${Q}$ </tex-math></inline-formula> filtering to enhance the resolution of seismic data. Since the attenuation process depends on time and frequency, it is routinely performed in the time–frequency domain. The synchrosqueezing transform (SST), which provides highly localized time–frequency representations for the nonstationary signals due to reduced spectral smearing, is applied to implement the inverse <inline-formula> <tex-math notation=\"LaTeX\">${Q}$ </tex-math></inline-formula> filtering scheme. However, the amplitude compensation process is unstable because energy amplification is involved. To stabilize it, the amplitude compensation is regarded as an inverse problem with an L1-norm regularization term in the SST domain. The iteratively reweighted least-squares algorithm is used to solve the regularized inverse problem. Synthetic and real data examples illustrate the stability and effectiveness of the proposed method.","PeriodicalId":13046,"journal":{"name":"IEEE Geoscience and Remote Sensing Letters","volume":"15 1","pages":"132-136"},"PeriodicalIF":4.8,"publicationDate":"2018-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1109/LGRS.2017.2777598","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"62472023","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-01-01DOI: 10.1109/LGRS.2017.2778421
Avadh Bihari Narayan, A. Tiwari, R. Dwivedi, O. Dikshit
Persistent scatterer (PS) pixels contain highly coherent information, which is used in the estimation of geophysical parameters of interest. Conventionally, PS pixels are selected on the basis of the estimated noise present in the spatially uncorrelated phase component along with look-angle error. The phase history of selected PS pixels is corrected for the look-angle error followed by phase unwrapping and extraction of spatially correlated nuisance phase component leading to displacement estimation. In this letter, a novel PS selection method, which is based on a new index called the similar time-series interferometric pixels (STIPs) representing the number of neighborhood pixels with similar phase history, is proposed. In this approach, apart from PS selection, corresponding set of STIP is also used in refining look-angle error estimation. The efficiency of the proposed InSAR processing chain is demonstrated for the Sentinel-1A single look complex images of Rajmahal, Jharkhand, India, predominantly a coal mines area. Results, when compared with the conventional PS processing technique, reveal substantial improvement in terms of extracting more number of reliable PS with enhanced density.
{"title":"Persistent Scatter Identification and Look-Angle Error Estimation Using Similar Time-Series Interferometric Pixels","authors":"Avadh Bihari Narayan, A. Tiwari, R. Dwivedi, O. Dikshit","doi":"10.1109/LGRS.2017.2778421","DOIUrl":"https://doi.org/10.1109/LGRS.2017.2778421","url":null,"abstract":"Persistent scatterer (PS) pixels contain highly coherent information, which is used in the estimation of geophysical parameters of interest. Conventionally, PS pixels are selected on the basis of the estimated noise present in the spatially uncorrelated phase component along with look-angle error. The phase history of selected PS pixels is corrected for the look-angle error followed by phase unwrapping and extraction of spatially correlated nuisance phase component leading to displacement estimation. In this letter, a novel PS selection method, which is based on a new index called the similar time-series interferometric pixels (STIPs) representing the number of neighborhood pixels with similar phase history, is proposed. In this approach, apart from PS selection, corresponding set of STIP is also used in refining look-angle error estimation. The efficiency of the proposed InSAR processing chain is demonstrated for the Sentinel-1A single look complex images of Rajmahal, Jharkhand, India, predominantly a coal mines area. Results, when compared with the conventional PS processing technique, reveal substantial improvement in terms of extracting more number of reliable PS with enhanced density.","PeriodicalId":13046,"journal":{"name":"IEEE Geoscience and Remote Sensing Letters","volume":"15 1","pages":"147-150"},"PeriodicalIF":4.8,"publicationDate":"2018-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1109/LGRS.2017.2778421","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"62472131","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-01-01DOI: 10.1109/LGRS.2017.2779040
Han Ma, S. Liang, Zhiqiang Xiao, Dongdong Wang
Traditional methods for estimating land-surface parameters from remotely sensed data generally focus on a single parameter with a specific spectral region, resulting in physical and spatiotemporal inconsistencies in current satellite products. We recently proposed a unified inversion scheme to estimate a suite of parameters simultaneously from both visible and near-infrared and thermal-infrared MODIS data. In this letter, we implemented this scheme to estimate six time-series parameters [leaf area index, fraction of absorbed photosynthetically active radiation, surface albedo, land-surface emissivity, land-surface temperature (LST), and upwelling longwave radiation (LWUP)] from the Visible Infrared Imaging Radiometer Suite (VIIRS) data. Several components of these schemes are refined, including the incorporation of a snow bidirectional reflectance distribution function model, determination of the best band combination, and better estimation of the snow-covered surface emissivity by accounting for the snow-cover fraction. Validation using the measurements at 12 sites of SURFRAD, CarboEuropeIP, and FLUXNET, and intercomparisons with MODIS and Global Land-Surface Satellite products, are carried out: the retrieved albedo, LST, and LWUP achieved accuracies ( $R^{2}$ ) of 0.77, 0.96, and 0.95, root mean square errors of 0.06, 2.9 K, and 18.3 W/m2, and biases of 0.01, 0.09 K, and −0.08 W/m2, respectively. The retrieved parameters can achieve comparable or higher accuracy than existing products, which indicates that the unified algorithm can be applied effectively to the VIIRS data with high physical and temporal consistency and accuracy.
{"title":"Simultaneous Estimation of Multiple Land-Surface Parameters From VIIRS Optical-Thermal Data","authors":"Han Ma, S. Liang, Zhiqiang Xiao, Dongdong Wang","doi":"10.1109/LGRS.2017.2779040","DOIUrl":"https://doi.org/10.1109/LGRS.2017.2779040","url":null,"abstract":"Traditional methods for estimating land-surface parameters from remotely sensed data generally focus on a single parameter with a specific spectral region, resulting in physical and spatiotemporal inconsistencies in current satellite products. We recently proposed a unified inversion scheme to estimate a suite of parameters simultaneously from both visible and near-infrared and thermal-infrared MODIS data. In this letter, we implemented this scheme to estimate six time-series parameters [leaf area index, fraction of absorbed photosynthetically active radiation, surface albedo, land-surface emissivity, land-surface temperature (LST), and upwelling longwave radiation (LWUP)] from the Visible Infrared Imaging Radiometer Suite (VIIRS) data. Several components of these schemes are refined, including the incorporation of a snow bidirectional reflectance distribution function model, determination of the best band combination, and better estimation of the snow-covered surface emissivity by accounting for the snow-cover fraction. Validation using the measurements at 12 sites of SURFRAD, CarboEuropeIP, and FLUXNET, and intercomparisons with MODIS and Global Land-Surface Satellite products, are carried out: the retrieved albedo, LST, and LWUP achieved accuracies ( $R^{2}$ ) of 0.77, 0.96, and 0.95, root mean square errors of 0.06, 2.9 K, and 18.3 W/m2, and biases of 0.01, 0.09 K, and −0.08 W/m2, respectively. The retrieved parameters can achieve comparable or higher accuracy than existing products, which indicates that the unified algorithm can be applied effectively to the VIIRS data with high physical and temporal consistency and accuracy.","PeriodicalId":13046,"journal":{"name":"IEEE Geoscience and Remote Sensing Letters","volume":"15 1","pages":"156-160"},"PeriodicalIF":4.8,"publicationDate":"2018-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1109/LGRS.2017.2779040","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"62472164","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-01-01DOI: 10.1109/LGRS.2017.2776113
Jinghui Yang, Jinxi Qian
In this letter, a multiscale joint collaborative representation with locally adaptive dictionary (MLJCRC) method is proposed for hyperspectral image classification. Based on the joint collaborative representation model, instead of selecting only a single region scale, MLJCRC incorporates complementary contextual information into classification by multiplying different scales with distinct spatial structures and characteristics. Also, MLJCRC uses a locally adaptive dictionary to reduce the influence of irrelevant pixels on representation, which improves the classification accuracy. The results of experiments on Indian Pines data and Pavia University data demonstrate that the proposed method performs better than support vector machine, sparse representation classification, and other collaborative representation-based classifications.
{"title":"Hyperspectral Image Classification via Multiscale Joint Collaborative Representation With Locally Adaptive Dictionary","authors":"Jinghui Yang, Jinxi Qian","doi":"10.1109/LGRS.2017.2776113","DOIUrl":"https://doi.org/10.1109/LGRS.2017.2776113","url":null,"abstract":"In this letter, a multiscale joint collaborative representation with locally adaptive dictionary (MLJCRC) method is proposed for hyperspectral image classification. Based on the joint collaborative representation model, instead of selecting only a single region scale, MLJCRC incorporates complementary contextual information into classification by multiplying different scales with distinct spatial structures and characteristics. Also, MLJCRC uses a locally adaptive dictionary to reduce the influence of irrelevant pixels on representation, which improves the classification accuracy. The results of experiments on Indian Pines data and Pavia University data demonstrate that the proposed method performs better than support vector machine, sparse representation classification, and other collaborative representation-based classifications.","PeriodicalId":13046,"journal":{"name":"IEEE Geoscience and Remote Sensing Letters","volume":"15 1","pages":"112-116"},"PeriodicalIF":4.8,"publicationDate":"2018-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1109/LGRS.2017.2776113","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"62471930","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2018-01-01DOI: 10.1109/LGRS.2017.2777264
F. Biondi
The problem in obtaining stable motion estimation of maritime targets is that sea clutter makes wake structure detection and reconnaissance difficult. This letter presents a complete procedure for the automatic estimation of maritime target motion parameters by evaluating the generated Kelvin waves detected in synthetic aperture radar (SAR) images. The algorithm consists in evaluating a dual-stage low-rank plus sparse decomposition (LRSD) assisted by Radon transform (RT) for clutter reduction, sparse object detection, precise wake inclination estimation, and Kelvin wave spectral analysis. The algorithm is based on the robust principal component analysis (RPCA) implemented by convex programming. The LRSD algorithm permits the extrapolation of sparse objects of interest consisting of the maritime targets and the Kelvin pattern from the unchanging low-rank background. This dual-stage RPCA and RT applied to SAR surveillance permits fast detection and enhanced motion parameter estimation of maritime targets.
{"title":"Low-Rank Plus Sparse Decomposition and Localized Radon Transform for Ship-Wake Detection in Synthetic Aperture Radar Images","authors":"F. Biondi","doi":"10.1109/LGRS.2017.2777264","DOIUrl":"https://doi.org/10.1109/LGRS.2017.2777264","url":null,"abstract":"The problem in obtaining stable motion estimation of maritime targets is that sea clutter makes wake structure detection and reconnaissance difficult. This letter presents a complete procedure for the automatic estimation of maritime target motion parameters by evaluating the generated Kelvin waves detected in synthetic aperture radar (SAR) images. The algorithm consists in evaluating a dual-stage low-rank plus sparse decomposition (LRSD) assisted by Radon transform (RT) for clutter reduction, sparse object detection, precise wake inclination estimation, and Kelvin wave spectral analysis. The algorithm is based on the robust principal component analysis (RPCA) implemented by convex programming. The LRSD algorithm permits the extrapolation of sparse objects of interest consisting of the maritime targets and the Kelvin pattern from the unchanging low-rank background. This dual-stage RPCA and RT applied to SAR surveillance permits fast detection and enhanced motion parameter estimation of maritime targets.","PeriodicalId":13046,"journal":{"name":"IEEE Geoscience and Remote Sensing Letters","volume":"15 1","pages":"117-121"},"PeriodicalIF":4.8,"publicationDate":"2018-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1109/LGRS.2017.2777264","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"62471938","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-12-01DOI: 10.1109/LGRS.2017.2778749
D. Ratha, A. Bhattacharya, A. Frery
In this letter, we propose a novel technique for obtaining scattering components from polarimetric synthetic aperture radar (PolSAR) data using the geodesic distance on the unit sphere. This geodesic distance is obtained between an elementary target and the observed Kennaugh matrix, and it is further utilized to compute a similarity measure between scattering mechanisms. The normalized similarity measure for each elementary target is then modulated with the total scattering power (Span). This measure is used to categorize pixels into three categories, i.e., odd-bounce, double-bounce, and volume, depending on which of the above scattering mechanisms dominate. Then the maximum likelihood classifier of Lee et al. based on the complex Wishart distribution is iteratively used for each category. Dominant scattering mechanisms are thus preserved in this classification scheme. We show results for L-band AIRSAR and ALOS-2 data sets acquired over San Francisco and Mumbai, respectively. The scattering mechanisms are better preserved using the proposed methodology than the unsupervised classification results using the Freeman–Durden scattering powers on an orientation angle corrected PolSAR image. Furthermore: 1) the scattering similarity is a completely nonnegative quantity unlike the negative powers that might occur in double-bounce and odd-bounce scattering component under Freeman–Durden decomposition and 2) the methodology can be extended to more canonical targets as well as for bistatic scattering.
{"title":"Unsupervised Classification of PolSAR Data Using a Scattering Similarity Measure Derived From a Geodesic Distance","authors":"D. Ratha, A. Bhattacharya, A. Frery","doi":"10.1109/LGRS.2017.2778749","DOIUrl":"https://doi.org/10.1109/LGRS.2017.2778749","url":null,"abstract":"In this letter, we propose a novel technique for obtaining scattering components from polarimetric synthetic aperture radar (PolSAR) data using the geodesic distance on the unit sphere. This geodesic distance is obtained between an elementary target and the observed Kennaugh matrix, and it is further utilized to compute a similarity measure between scattering mechanisms. The normalized similarity measure for each elementary target is then modulated with the total scattering power (Span). This measure is used to categorize pixels into three categories, i.e., odd-bounce, double-bounce, and volume, depending on which of the above scattering mechanisms dominate. Then the maximum likelihood classifier of Lee et al. based on the complex Wishart distribution is iteratively used for each category. Dominant scattering mechanisms are thus preserved in this classification scheme. We show results for L-band AIRSAR and ALOS-2 data sets acquired over San Francisco and Mumbai, respectively. The scattering mechanisms are better preserved using the proposed methodology than the unsupervised classification results using the Freeman–Durden scattering powers on an orientation angle corrected PolSAR image. Furthermore: 1) the scattering similarity is a completely nonnegative quantity unlike the negative powers that might occur in double-bounce and odd-bounce scattering component under Freeman–Durden decomposition and 2) the methodology can be extended to more canonical targets as well as for bistatic scattering.","PeriodicalId":13046,"journal":{"name":"IEEE Geoscience and Remote Sensing Letters","volume":"15 1","pages":"151-155"},"PeriodicalIF":4.8,"publicationDate":"2017-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1109/LGRS.2017.2778749","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"44055107","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-10-01DOI: 10.1109/LGRS.2017.2729940
Weiwei Sun, Chun Liu, Yan Xu, Long Tian, Weiyue Li
A band-weighted support vector machine (BWSVM) method is proposed to classify hyperspectral imagery (HSI). The BWSVM presents an L1 penalty term of band weight vector to regularize the regular SVM model. The L1 norm regularization term guarantees the sparsity of band weights and describes potentially divergent contributions from different bands in modeling the binary SVM model. The BWSVM adopts the KerNel iterative feature extraction algorithm to minimize the nonconvex program. It linearizes nonlinear kernels and iteratively optimizes two convex subproblems with respect to both sample coefficients and band weights. The class label is determined by picking the largest sample coefficients from all its binary models of BWSVM. Two popular HSI data sets are utilized to testify the classification performance of BWSVM. Experimental results show that the BWSVM outperforms three state-of-the-art classifiers including SVM, random forest, and k-nearest neighbor.
{"title":"A Band-Weighted Support Vector Machine Method for Hyperspectral Imagery Classification","authors":"Weiwei Sun, Chun Liu, Yan Xu, Long Tian, Weiyue Li","doi":"10.1109/LGRS.2017.2729940","DOIUrl":"https://doi.org/10.1109/LGRS.2017.2729940","url":null,"abstract":"A band-weighted support vector machine (BWSVM) method is proposed to classify hyperspectral imagery (HSI). The BWSVM presents an L1 penalty term of band weight vector to regularize the regular SVM model. The L1 norm regularization term guarantees the sparsity of band weights and describes potentially divergent contributions from different bands in modeling the binary SVM model. The BWSVM adopts the KerNel iterative feature extraction algorithm to minimize the nonconvex program. It linearizes nonlinear kernels and iteratively optimizes two convex subproblems with respect to both sample coefficients and band weights. The class label is determined by picking the largest sample coefficients from all its binary models of BWSVM. Two popular HSI data sets are utilized to testify the classification performance of BWSVM. Experimental results show that the BWSVM outperforms three state-of-the-art classifiers including SVM, random forest, and k-nearest neighbor.","PeriodicalId":13046,"journal":{"name":"IEEE Geoscience and Remote Sensing Letters","volume":"14 1","pages":"1710-1714"},"PeriodicalIF":4.8,"publicationDate":"2017-10-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1109/LGRS.2017.2729940","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"44951850","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","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：