Pub Date : 2003-10-27DOI: 10.1109/WARSD.2003.1295197
Sanghun Lim, V. Chandrasekar
Hydrometeor classification system using fuzzy logic technique based on dual-polarization radar measurements is presented. In this study, five radar measurements (horizontal reflectivity, differential reflectivity, specific differential phase, correlation coefficient, and linear depolarization ratio), and height relating to environmental melting level are used as input variables of the system. The hydrometeor classification system chooses one of nine different hydrometeor categories as output. The system presented in this paper is a further development of an existing hydrometeor classification system model developed at Colorado State University. The hydrometeor classification system is evaluated by comparison against the in situ sample data collected by instrumentation on T-28 aircraft during Severe Thunderstorm Electrification and Precipitation Study (STEPS).
{"title":"Hydrometeor classification system using dual-polarization radar measurements","authors":"Sanghun Lim, V. Chandrasekar","doi":"10.1109/WARSD.2003.1295197","DOIUrl":"https://doi.org/10.1109/WARSD.2003.1295197","url":null,"abstract":"Hydrometeor classification system using fuzzy logic technique based on dual-polarization radar measurements is presented. In this study, five radar measurements (horizontal reflectivity, differential reflectivity, specific differential phase, correlation coefficient, and linear depolarization ratio), and height relating to environmental melting level are used as input variables of the system. The hydrometeor classification system chooses one of nine different hydrometeor categories as output. The system presented in this paper is a further development of an existing hydrometeor classification system model developed at Colorado State University. The hydrometeor classification system is evaluated by comparison against the in situ sample data collected by instrumentation on T-28 aircraft during Severe Thunderstorm Electrification and Precipitation Study (STEPS).","PeriodicalId":395735,"journal":{"name":"IEEE Workshop on Advances in Techniques for Analysis of Remotely Sensed Data, 2003","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2003-10-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125002960","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 : 2003-10-27DOI: 10.1109/WARSD.2003.1295215
J. A. Gualtieri
Two different areas of current research in hyperspectral remote sensing are addressed: (1) supervised learning using all the hyperspectral bands as based on the recently introduced method called the support vector machine. (2) Hyperspectral remote sensing in shallow water to retrieve benthic properties including depth and albedo on the sea floor. The support vector technique is applied to land agricultural scenes acquired by AVIRIS with up to 16 classes, and is shown to give improved results over a number of methods all applied to the same scene. Hyperspectral remote sensing in shallow water is demonstrated on an AVIRIS scene acquired in Kaneohe Bay Hawaii, where reasonable depths and bottom albedos are retrieved. The method is based on physical modeling of the propagation of light though the atmosphere and physical modeling of the propagation of light through the water column above the sea floor. The results for shallow water remote sensing are extended by a physically realistic simulation using AVIRIS at-sensor data to model cases of spatial resolution and signal to noise ratios that might exist for a hyperspectral sensor in geostationary orbit.
{"title":"Hyperspectral analysis, the support vector machine, and land and benthic habitats","authors":"J. A. Gualtieri","doi":"10.1109/WARSD.2003.1295215","DOIUrl":"https://doi.org/10.1109/WARSD.2003.1295215","url":null,"abstract":"Two different areas of current research in hyperspectral remote sensing are addressed: (1) supervised learning using all the hyperspectral bands as based on the recently introduced method called the support vector machine. (2) Hyperspectral remote sensing in shallow water to retrieve benthic properties including depth and albedo on the sea floor. The support vector technique is applied to land agricultural scenes acquired by AVIRIS with up to 16 classes, and is shown to give improved results over a number of methods all applied to the same scene. Hyperspectral remote sensing in shallow water is demonstrated on an AVIRIS scene acquired in Kaneohe Bay Hawaii, where reasonable depths and bottom albedos are retrieved. The method is based on physical modeling of the propagation of light though the atmosphere and physical modeling of the propagation of light through the water column above the sea floor. The results for shallow water remote sensing are extended by a physically realistic simulation using AVIRIS at-sensor data to model cases of spatial resolution and signal to noise ratios that might exist for a hyperspectral sensor in geostationary orbit.","PeriodicalId":395735,"journal":{"name":"IEEE Workshop on Advances in Techniques for Analysis of Remotely Sensed Data, 2003","volume":"235 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2003-10-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121144299","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 : 2003-10-27DOI: 10.1109/WARSD.2003.1295214
C. A. Shah, P. Watanachaturaporn, P. Varshney, Manoj K. Arora
In this paper, we present a summary of our ongoing research on the classification of hyperspectral images. We are experimenting with both supervised and unsupervised algorithms. In particular, we have developed an unsupervised classification algorithm based on Independent Component Analysis (ICA). This algorithm is known as the ICA mixture model (ICAMM) algorithm and has shown promising results. In addition, we are investigating the use of Support Vector Machines (SVMs), a supervised approach for the classification of hyperspectral data. We have employed the Lagrangian optimization method and call our classifier the Lagrangian SVM (LSVM) classifier. Classification accuracy of these classifiers has been assessed using an error matrix based overall accuracy measure.
{"title":"Some recent results on hyperspectral image classification","authors":"C. A. Shah, P. Watanachaturaporn, P. Varshney, Manoj K. Arora","doi":"10.1109/WARSD.2003.1295214","DOIUrl":"https://doi.org/10.1109/WARSD.2003.1295214","url":null,"abstract":"In this paper, we present a summary of our ongoing research on the classification of hyperspectral images. We are experimenting with both supervised and unsupervised algorithms. In particular, we have developed an unsupervised classification algorithm based on Independent Component Analysis (ICA). This algorithm is known as the ICA mixture model (ICAMM) algorithm and has shown promising results. In addition, we are investigating the use of Support Vector Machines (SVMs), a supervised approach for the classification of hyperspectral data. We have employed the Lagrangian optimization method and call our classifier the Lagrangian SVM (LSVM) classifier. Classification accuracy of these classifiers has been assessed using an error matrix based overall accuracy measure.","PeriodicalId":395735,"journal":{"name":"IEEE Workshop on Advances in Techniques for Analysis of Remotely Sensed Data, 2003","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2003-10-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129400513","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 : 2003-10-27DOI: 10.1109/WARSD.2003.1295176
Chulhee Lee, E. Choi
We propose compression algorithms for hyperspectral images with enhanced discriminant features. As the dimension of remotely sensed images increases, the need for efficient compression algorithms for hyperspectral images also increases. However, when hyperspectral images are compressed with conventional image compression algorithms, which have been developed to minimize mean squared errors, discriminant features of the original data may be lost during the compression process. In this paper, we propose to apply preprocessing prior to compression in order to preserve such discriminant information. In particular, we enhance discriminant features before a compression algorithm is applied. Experiments show that the proposed method provides improved classification accuracies than the existing compression algorithms.
{"title":"Compression of hyperspectral images with enhanced discriminant features","authors":"Chulhee Lee, E. Choi","doi":"10.1109/WARSD.2003.1295176","DOIUrl":"https://doi.org/10.1109/WARSD.2003.1295176","url":null,"abstract":"We propose compression algorithms for hyperspectral images with enhanced discriminant features. As the dimension of remotely sensed images increases, the need for efficient compression algorithms for hyperspectral images also increases. However, when hyperspectral images are compressed with conventional image compression algorithms, which have been developed to minimize mean squared errors, discriminant features of the original data may be lost during the compression process. In this paper, we propose to apply preprocessing prior to compression in order to preserve such discriminant information. In particular, we enhance discriminant features before a compression algorithm is applied. Experiments show that the proposed method provides improved classification accuracies than the existing compression algorithms.","PeriodicalId":395735,"journal":{"name":"IEEE Workshop on Advances in Techniques for Analysis of Remotely Sensed Data, 2003","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2003-10-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129050043","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 : 2003-10-27DOI: 10.1109/WARSD.2003.1295199
H. Ren, Q. Du, Chein-I. Chang, J. Jensen
Constrained Energy Minimization (CEM) has been widely used for target detection in hyperspectral remote sensing imagery. It detects the desired target signal source using a unity constraint while suppressing noise and unknown signal sources by minimizing the average output power. Base on the design CEM can only detect one target source at a time. In order to simultaneously detect multiple targets in a single image, several approaches are developed, including Multiple-Target CEM (MTCEM), Sum CEM (SCEM) and Winner-Take-All CEM (WTACEM). Interestingly, the sensitivity of noise and interference seems to play a role in the detection performance. Unfortunately, this issue has not been investigated. In this paper, we take up this problem and conduct a quantitative study of the noise and interference suppression abilities of LCMV, SCEM, WTACEM for multiple-target detection.
约束能量最小化(CEM)在高光谱遥感图像目标检测中得到了广泛的应用。它使用统一约束检测期望的目标信号源，同时通过最小化平均输出功率来抑制噪声和未知信号源。基于这种设计，CEM一次只能检测一个目标源。为了在单幅图像中同时检测多个目标，发展了多目标CEM (MTCEM)、Sum CEM (SCEM)和赢者通吃CEM (WTACEM)等方法。有趣的是，噪声和干扰的灵敏度似乎在检测性能中起作用。不幸的是，这个问题还没有得到调查。本文针对这一问题，对LCMV、SCEM、WTACEM在多目标检测中的噪声和干扰抑制能力进行了定量研究。
{"title":"Comparison between constrained energy minimization based approaches for hyperspectral imagery","authors":"H. Ren, Q. Du, Chein-I. Chang, J. Jensen","doi":"10.1109/WARSD.2003.1295199","DOIUrl":"https://doi.org/10.1109/WARSD.2003.1295199","url":null,"abstract":"Constrained Energy Minimization (CEM) has been widely used for target detection in hyperspectral remote sensing imagery. It detects the desired target signal source using a unity constraint while suppressing noise and unknown signal sources by minimizing the average output power. Base on the design CEM can only detect one target source at a time. In order to simultaneously detect multiple targets in a single image, several approaches are developed, including Multiple-Target CEM (MTCEM), Sum CEM (SCEM) and Winner-Take-All CEM (WTACEM). Interestingly, the sensitivity of noise and interference seems to play a role in the detection performance. Unfortunately, this issue has not been investigated. In this paper, we take up this problem and conduct a quantitative study of the noise and interference suppression abilities of LCMV, SCEM, WTACEM for multiple-target detection.","PeriodicalId":395735,"journal":{"name":"IEEE Workshop on Advances in Techniques for Analysis of Remotely Sensed Data, 2003","volume":"472 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2003-10-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124391079","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 : 2003-10-27DOI: 10.1109/WARSD.2003.1295184
W.W. Stoner
The convex hull methods for estimating spectral endmembers are subject to bias errors: mixed pixel bias - if all of the available pixels are mosaics of all m endmembers, the convex-hull derived endmember spectra are biased towards the centroid of the true endmember spectra; noise bias - additive Gaussian measurement noise inflates the convex hull away from the centroid of the noise-free convex hull. The noise bias error grows with the pixel count. This vulnerability to mixed pixel bias and noise bias prompts the following questions. Does the convex hull method throw away information by discarding the pixels lying inside the convex hull? Can bias error estimates be developed for convex-hull derived endmembers? Can bias-resistant endmember estimation methods be found? What is the gain in accuracy of the endmember estimates with increasing pixel count? What is the gain in accuracy with increasing density of pixels in the n-dimensional neighborhood of the true endmember? The following analysis focuses on these questions by omitting all sources of noise and distortion except the number and distribution of the samples in the neighborhood of the endmember.
{"title":"Towards a statistical error estimate for convex-hull derived endmembers","authors":"W.W. Stoner","doi":"10.1109/WARSD.2003.1295184","DOIUrl":"https://doi.org/10.1109/WARSD.2003.1295184","url":null,"abstract":"The convex hull methods for estimating spectral endmembers are subject to bias errors: mixed pixel bias - if all of the available pixels are mosaics of all m endmembers, the convex-hull derived endmember spectra are biased towards the centroid of the true endmember spectra; noise bias - additive Gaussian measurement noise inflates the convex hull away from the centroid of the noise-free convex hull. The noise bias error grows with the pixel count. This vulnerability to mixed pixel bias and noise bias prompts the following questions. Does the convex hull method throw away information by discarding the pixels lying inside the convex hull? Can bias error estimates be developed for convex-hull derived endmembers? Can bias-resistant endmember estimation methods be found? What is the gain in accuracy of the endmember estimates with increasing pixel count? What is the gain in accuracy with increasing density of pixels in the n-dimensional neighborhood of the true endmember? The following analysis focuses on these questions by omitting all sources of noise and distortion except the number and distribution of the samples in the neighborhood of the endmember.","PeriodicalId":395735,"journal":{"name":"IEEE Workshop on Advances in Techniques for Analysis of Remotely Sensed Data, 2003","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2003-10-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130812386","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 : 2003-10-27DOI: 10.1109/WARSD.2003.1295193
S. Chettri, W. Campbell
This paper applies two recent methods to denoise remotely sensed images - wavelet based Markov Random Field (MRF) methods, Independent Component Analysis (ICA) and compares them with the standard Wiener filter. In order to facilitate the continued use of these methods in remote sensing the theory behind each method is discussed in detail. Subsequently they are applied to de-noising remotely sensed images. The efficiency of each algorithm is obtained by computing the signal to noise ratio (SNR) before and after de-noising. Results indicate that the MRF based methods, though slightly more complicated to program and only marginally slower than ICA de-noising, generally perform better than both ICA and Wiener filtering. The article ends by discussing future areas of research in de-noising remotely sensed images.
{"title":"De-noising remotely sensed digital imagery","authors":"S. Chettri, W. Campbell","doi":"10.1109/WARSD.2003.1295193","DOIUrl":"https://doi.org/10.1109/WARSD.2003.1295193","url":null,"abstract":"This paper applies two recent methods to denoise remotely sensed images - wavelet based Markov Random Field (MRF) methods, Independent Component Analysis (ICA) and compares them with the standard Wiener filter. In order to facilitate the continued use of these methods in remote sensing the theory behind each method is discussed in detail. Subsequently they are applied to de-noising remotely sensed images. The efficiency of each algorithm is obtained by computing the signal to noise ratio (SNR) before and after de-noising. Results indicate that the MRF based methods, though slightly more complicated to program and only marginally slower than ICA de-noising, generally perform better than both ICA and Wiener filtering. The article ends by discussing future areas of research in de-noising remotely sensed images.","PeriodicalId":395735,"journal":{"name":"IEEE Workshop on Advances in Techniques for Analysis of Remotely Sensed Data, 2003","volume":"15 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2003-10-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130973441","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 : 2003-10-27DOI: 10.1109/WARSD.2003.1295224
L. Bruzzone, L. Carlin, F. Melgani
This paper presents a novel residual-based approach to classification of remote sensing images. The proposed approach aims at increasing the accuracy of classification methods explicitly (or implicitly) inspired to the Bayesian decision theory. In particular, an architecture composed of an ensemble of estimators is used in order to estimate the residual errors in the class conditional posterior probabilities estimated by the Bayesian classifier considered. In order to avoid overfitting of the training data, a technique based on the analysis of class conditional entropy measures of omission and commission errors is used for adaptively evaluating the number of estimators to be included in the ensemble. Experimental results obtained on two multisource and multisensor data sets (characterized by different complexities) confirm the effectiveness of the proposed system.
{"title":"A residual-based approach to classification of remote sensing images","authors":"L. Bruzzone, L. Carlin, F. Melgani","doi":"10.1109/WARSD.2003.1295224","DOIUrl":"https://doi.org/10.1109/WARSD.2003.1295224","url":null,"abstract":"This paper presents a novel residual-based approach to classification of remote sensing images. The proposed approach aims at increasing the accuracy of classification methods explicitly (or implicitly) inspired to the Bayesian decision theory. In particular, an architecture composed of an ensemble of estimators is used in order to estimate the residual errors in the class conditional posterior probabilities estimated by the Bayesian classifier considered. In order to avoid overfitting of the training data, a technique based on the analysis of class conditional entropy measures of omission and commission errors is used for adaptively evaluating the number of estimators to be included in the ensemble. Experimental results obtained on two multisource and multisensor data sets (characterized by different complexities) confirm the effectiveness of the proposed system.","PeriodicalId":395735,"journal":{"name":"IEEE Workshop on Advances in Techniques for Analysis of Remotely Sensed Data, 2003","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2003-10-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114748104","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 : 2003-10-27DOI: 10.1109/WARSD.2003.1295202
S.B. Ziegeler, H. Tamhankar, J. Fowler, L. Bruce
Watermarking is widely being explored as a means of providing protection of ownership rights for multimedia data, and there has been increasing interest in applying watermarking to remotely sensed data for this same purpose. However, watermarking techniques developed for multimedia cannot be applied directly to remotely sensed data due the fact that the analytic integrity of the data, rather than perceptual quality, is of primary importance. In this paper, a watermarking technique for remotely sensed data based on the discrete wavelet transform (DWT) is proposed, and its impact on unsupervised classification as well as attacks such as cropping is studied.
{"title":"Wavelet-based watermarking of remotely sensed imagery tailored to classification performance","authors":"S.B. Ziegeler, H. Tamhankar, J. Fowler, L. Bruce","doi":"10.1109/WARSD.2003.1295202","DOIUrl":"https://doi.org/10.1109/WARSD.2003.1295202","url":null,"abstract":"Watermarking is widely being explored as a means of providing protection of ownership rights for multimedia data, and there has been increasing interest in applying watermarking to remotely sensed data for this same purpose. However, watermarking techniques developed for multimedia cannot be applied directly to remotely sensed data due the fact that the analytic integrity of the data, rather than perceptual quality, is of primary importance. In this paper, a watermarking technique for remotely sensed data based on the discrete wavelet transform (DWT) is proposed, and its impact on unsupervised classification as well as attacks such as cropping is studied.","PeriodicalId":395735,"journal":{"name":"IEEE Workshop on Advances in Techniques for Analysis of Remotely Sensed Data, 2003","volume":"78 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2003-10-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121115666","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 : 2003-10-27DOI: 10.1109/WARSD.2003.1295187
Jiang Li, L. Bruce
Spectral unmixing is a quantitative analysis procedure used to recognize constituent ground cover materials (or endmembers) and obtain their mixing proportions (or abundances) from a mixed pixel. The endmember abundances may be estimated using the least squares estimation (LSE) method based on the linear mixture model. This paper investigates the use of spectral dimensionality reduction as a preprocessing tool for hyperspectral linear unmixing. Four dimensionality reduction methods are investigated and compared; these include methods based on the discrete wavelet transform (DWT), discrete cosine transform, principal component transform, and linear discriminant transform (LDT). Three sets of experiments are designed and implemented for evaluating the effects of the dimensionality reduction techniques on the LSE of endmember abundances. Experimental results show that the use of the DWT and LDT-based features extracted from the original hyperspectral signals can greatly improve the abundance estimation of endmembers. On average with these methods, the root-mean-square of the abundance estimation error is reduced by 20%.
{"title":"Improving the accuracy of linear pixel unmixing via appropriate endmember dimensionality reduction","authors":"Jiang Li, L. Bruce","doi":"10.1109/WARSD.2003.1295187","DOIUrl":"https://doi.org/10.1109/WARSD.2003.1295187","url":null,"abstract":"Spectral unmixing is a quantitative analysis procedure used to recognize constituent ground cover materials (or endmembers) and obtain their mixing proportions (or abundances) from a mixed pixel. The endmember abundances may be estimated using the least squares estimation (LSE) method based on the linear mixture model. This paper investigates the use of spectral dimensionality reduction as a preprocessing tool for hyperspectral linear unmixing. Four dimensionality reduction methods are investigated and compared; these include methods based on the discrete wavelet transform (DWT), discrete cosine transform, principal component transform, and linear discriminant transform (LDT). Three sets of experiments are designed and implemented for evaluating the effects of the dimensionality reduction techniques on the LSE of endmember abundances. Experimental results show that the use of the DWT and LDT-based features extracted from the original hyperspectral signals can greatly improve the abundance estimation of endmembers. On average with these methods, the root-mean-square of the abundance estimation error is reduced by 20%.","PeriodicalId":395735,"journal":{"name":"IEEE Workshop on Advances in Techniques for Analysis of Remotely Sensed Data, 2003","volume":"189 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2003-10-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122294630","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：