Spectral and Spatial Classification of Hyperspectral Data Using SVMs and Morphological Profiles
IEEE Transactions on Geoscience and Remote SensingPublished 1 November 2008Open access
Mathieu Fauvel, Jón Atli Benediktsson, Jón Atli Benediktsson, Jóhannes R. Sveinsson
Citations961
Generate an AI Snapshot to get a quick, structured summary of this paper.
Study Snapshot
ObjectiveStudy objective
MethodsResearch methodology
PopulationPopulation studied
Sample sizeSample sizes
OutcomesStudy outcomes here
ResultsStudy results comes here
LimitationsResearch study limitations comes here
A concise AI-generated summary of the paper will appear here once you click Generate AI Snapshot.
TL;DR
A method is proposed for the classification of urban hyperspectral data with high spatial resolution based on the fusion of the morphological information and the original hyperspectral data, i.e., the two vectors of attributes are concatenated into one feature vector.
Abstract
Abstract—A method is proposed for the classification of urban hyperspectral data with high spatial resolution. The approach is an extension of previous approaches and uses both the spatial and spectral information for classification. One previous approach is based on using several principal components from the hyper-spectral data and building several morphological profiles. These profiles can be used all together in one extended morphological profile. A shortcoming of that approach is that it was primarily designed for classification of urban structures and it does not fully utilize the spectral information in the data. Similarly, the commonly used pixel-wise classification of hyperspectral data is solely based on the spectral content and lacks information on the structure of the features in the image. The proposed method overcomes these problems and is based on the fusion of the morphological information and the original hyperspectral data, i.e., the two vectors of attributes are concatenated into one feature vector. After a reduction of the dimensionality the final classification is achieved using a Support Vector Machines classifier. The proposed approach is tested in experiments on ROSIS data from urban areas. Significant improvements are achieved in terms of accuracies when compared to results obtained for approaches based on the use of morphological profiles based on PCs only and conventional spectral classification. For instance, with one data set, the overall accuracy is increased from 79 % to 83 % without any feature reduction and to 87 % with feature reduction. The proposed approach also shows excellent results with a limited training set. Index Terms—Data fusion, hyperspectral data, support vector machines, feature extraction, extended morphological profile, high spatial resolution. I.
Keywords
Earth and Planetary SciencesEngineeringEnvironmental Science
Data Mining and Knowledge DiscoveryA Tutorial on Support Vector Machines for Pattern Recognition
16,433 Citations1998Christopher J. C. Burges
There are several arguments which support the observed high accuracy of SVMs, which are reviewed and numerous examples and proofs of most of the key theorems are given.
IEEE Transactions on Geoscience and Remote SensingClassification of hyperspectral remote sensing images with support vector machines
4,325 Citations2004Farid Melgani, Lorenzo Bruzzone
This paper addresses the problem of the classification of hyperspectral remote sensing images by support vector machines by understanding and assessing the potentialities of SVM classifiers in hyperdimensional feature spaces and concludes that SVMs are a valid and effective alternative to conventional pattern recognition approaches.
Remote Sensing Digital Image Analysis
1,518 Citations1999John A. Richards, Xiuping Jia
IEEE Transactions on Geoscience and Remote SensingClassification of hyperspectral data from urban areas based on extended morphological profiles
1,381 Citations2005Jón Atli Benediktsson, J.A. Palmason +1 more
A method based on mathematical morphology for preprocessing of the hyperspectral data is proposed, using opening and closing morphological transforms to isolate bright and dark structures in images, where bright/dark means brighter/darker than the surrounding features in the images.
IEEE Geoscience and Remote Sensing LettersComposite Kernels for Hyperspectral Image Classification
1,072 Citations2006Gustau Camps‐Valls, Luis Gómez‐Chova +3 more
This framework of composite kernels demonstrates enhanced classification accuracy as compared to traditional approaches that take into account the spectral information only, flexibility to balance between the spatial and spectral information in the classifier, and computational efficiency.
Photogrammetric Engineering & Remote SensingThematic Map Comparison
1,045 Citations2004Giles M. Foody
IEEE Transactions on Geoscience and Remote SensingA relative evaluation of multiclass image classification by support vector machines
1,022 Citations2004Giles M. Foody, Ajay Mathur
Although each classifier could yield a very accurate classification, > 90% correct, the classifiers differed in the ability to correctly label individual cases and so may be suitable candidates for an ensemble-based approach to classification.
Spectral and spatial classification of hyperspectral data using SVMs and morphological profiles
909 Citations2007Mathieu Fauvel, Jocelyn Chanussot +2 more
An approach has been proposed which is based on using several principal components from the hyperspectral data and build morphological profiles which can be used all together in one extended morphological profile for classification of urban structures.
IEEE Transactions on Geoscience and Remote SensingA new approach for the morphological segmentation of high-resolution satellite imagery
817 Citations2001Martino Pesaresi, Jón Atli Benediktsson
The proposed method performs well in the presence of both low radiometric contrast and relatively low spatial resolution, which may produce a textural effect, a border effect, and ambiguity in the object/background distinction.
IEEE Transactions on Geoscience and Remote SensingClassification and feature extraction for remote sensing images from urban areas based on morphological transformations
748 Citations2003Jón Atli Benediktsson, Martino Pesaresi +1 more
It is seen that relatively few features are needed to achieve the same classification accuracies as in the original feature space when classification of panchromatic high-resolution data from urban areas using morphological and neural approaches.
IEEE Transactions on Geoscience and Remote SensingThe effect of unlabeled samples in reducing the small sample size problem and mitigating the Hughes phenomenon
572 Citations1994B.M. Shahshahani, D. A. Landgrebe
By using additional unlabeled samples that are available at no extra cost, the performance may be improved, and therefore the Hughes phenomenon can be mitigated and therefore more representative estimates can be obtained.
IEEE Transactions on Geoscience and Remote SensingA Novel Transductive SVM for Semisupervised Classification of Remote-Sensing Images
560 Citations2006Lorenzo Bruzzone, Mingmin Chi +1 more
A novel modified TSVM classifier designed for addressing ill-posed remote-sensing problems is proposed that is able to mitigate the effects of suboptimal model selection and can address multiclass cases.
IEEE Transactions on Geoscience and Remote SensingToward an Optimal SVM Classification System for Hyperspectral Remote Sensing Images
427 Citations2006Yakoub Bazi, Farid Melgani
This paper proposes a classification system based on a genetic optimization framework formulated in such a way as to detect the best discriminative features without requiring the a priori setting of their number by the user and to estimate the best SVM parameters in a completely automatic way.
Proceedings of SPIE, the International Society for Optical Engineering/Proceedings of SPIE<title>Support vector machines for hyperspectral remote sensing classification</title>
422 Citations1999J. Anthony Gualtieri, Robert F. Cromp
A key feature of this classifier is its ability to use high-dimensional data without the usual recourse to a feature selection step to reduce the dimensionality of the data.
IEEE Transactions on Pattern Analysis and Machine IntelligenceFeature extraction based on decision boundaries
376 Citations1993C. Lee, D. A. Landgrebe
The proposed feature extraction algorithm has several desirable properties: it predicts the minimum number of features necessary to achieve the same classification accuracy as in the original space for a given pattern recognition problem; and it finds the necessary feature vectors.
IEEE Transactions on Geoscience and Remote SensingAdvances in mathematical morphology applied to geoscience and remote sensing
332 Citations2002Pierre Soille, Martino Pesaresi
It is shown that self-dual or even self-complementary filters are required in many applications to produce results independent of the local contrast of the searched image structures, and important issues overlooked in the past are addressed.
IEEE Transactions on Pattern Analysis and Machine IntelligenceCovariance matrix estimation and classification with limited training data
305 Citations1996Joseph Hoffbeck, D. A. Landgrebe
A new covariance matrix estimator useful for designing classifiers with limited training data is developed, and in experiments, this estimator achieved higher classification accuracy than the sample covariance matrices and common covariance Matrix estimates.
IEEE Geoscience and Remote Sensing LettersExploiting Spectral and Spatial Information in Hyperspectral Urban Data With High Resolution
260 Citations2004Fabio Dell’Acqua, Paolo Gamba +4 more
New methods for classification of hyperspectral remote sensing data are investigated, with the primary focus on multiple classifications and spatial analysis to improve mapping accuracy in urban areas.
IEEE Transactions on Geoscience and Remote SensingDecision Fusion for the Classification of Urban Remote Sensing Images
250 Citations2006Mathieu Fauvel, Jocelyn Chanussot +1 more
A general framework for combining information from several individual classifiers in multiclass classification is proposed, based on the definition of two measures of accuracy, which estimates for each pixel the reliability of the information provided by each classifier.
IEEE Transactions on Geoscience and Remote SensingAdaptive Bayesian contextual classification based on Markov random fields
210 Citations2002Q. Jackson, D. A. Landgrebe
IEEE Transactions on Geoscience and Remote SensingSemisupervised Classification of Hyperspectral Images by SVMs Optimized in the Primal
194 Citations2007Mingmin Chi, Lorenzo Bruzzone
This paper addresses classification of hyperspectral remote sensing images with kernel-based methods defined in the framework of semisupervised support vector machines by considering different (S3VMs) techniques that solve optimization directly in the primal formulation of the objective function.
IEEE Transactions on Geoscience and Remote SensingOne-Class Classification for Mapping a Specific Land-Cover Class: SVDD Classification of Fenland
161 Citations2007Carolina Sanchez‐Hernandez, Doreen S. Boyd +1 more
It is suggested that a one-class-classification approach could be appropriate when interest focuses on a specific class, as illustrated with the classification of fenland, a habitat of considerable conservation value, from Landsat Enhanced Thematic Mapper Plus imagery.
Support vector machines for classification of hyperspectral data
152 Citations2002J. Anthony Gualtieri, S. Chettri
The support vector machine, recently introduced by Boser, Guyon, and Vapnik is useful in solving supervised classification in high dimensions and its application to high dimensional hyperspectral data taken from NASA's AVIRIS sensor and from a commercially available sensor called AISA.
IEEE Geoscience and Remote Sensing LettersClassification of Remote Sensing Images From Urban Areas Using a Fuzzy Possibilistic Model
119 Citations2006Jocelyn Chanussot, Jón Atli Benediktsson +1 more
An interpretation of the derivative morphological profile (DMP) obtained with a granulometric approach is presented in terms of a fuzzy measurement of the characteristic size and contrast of each structure.
Evaluation of Kernels for Multiclass Classification of Hyperspectral Remote Sensing Data
117 Citations2006Mathieu Fauvel, Jocelyn Chanussot +1 more
The experimental results show that SVMs used with the two kernels are appropriate for remote sensing classification problems.
IEEE Transactions on Geoscience and Remote SensingPartially Supervised Oil-Slick Detection by SAR Imagery Using Kernel Expansion
102 Citations2006Grégoire Mercier, Fanny Girard‐Ardhuin
A semisupervised oil-slick detection is proposed by using a kernel-based abnormal detection into the wavelet decomposition of a SAR image, which performs accurate detection with no consideration to signal stationarity nor to the presence of strong backscatters (such as a ship).
IEEE Transactions on Geoscience and Remote SensingSupervised segmentation of remote sensing images based on a tree-structured MRF model
95 Citations2005Giovanni Poggi, Giuseppe Scarpa +1 more
This work proposes to use the tree-structured Markov random field model, which describes a K-ary field by means of a sequence of binary MRFs, each one corresponding to a node in the tree, for supervised segmentation of remote sensing images.
Classification of hyperspectral data from urban areas using morpholgical preprocessing and independent component analysis
91 Citations2005J.A. Palmason, Jón Atli Benediktsson +2 more
This paper investigates the use of independent components instead of principal components in extended Morphological profiles, i.e., selected independent components are used as base images for an extended morphological profile and used as inputs to a neural network classifier.
IEEE Transactions on Geoscience and Remote SensingA robust classification procedure based on mixture classifiers and nonparametric weighted feature extraction
51 Citations2002Bor‐Chen Kuo, D. A. Landgrebe
The intent of this study is to design a classification procedure that is robust and maximally effective, but which provides the analyst with significant assists, thus simplifying the analyst's task.
IEEE Transactions on Geoscience and Remote SensingMorphological Texture Features for Unsupervised and Supervised Segmentations of Natural Landscapes
50 Citations2007Irene Epifanio, Pierre Soille
To segment high-resolution images of natural landscapes into different cover types, morphological texture features are used in order to avoid the limitations of spectral features.
Fusion of Morphological and Spectral Information for Classification of Hyperspectal Urban Remote Sensing Data
17 Citations2006J.A. Palmason, Jón Atli Benediktsson +2 more
An extension is proposed in this paper in order to overcome the problems with the extended morphological profile approach and achieve significant improvements in terms of accuracies when compared to results of approaches based on the use of morphological profiles based on PCs only and conventional statistical approaches.
Performance analysis and comparison of neural networks and support vector machines classifier
11 Citations2004Enhui Zheng, Ping Li +1 more
The theory foundation and classification algorithm of neural networks (NN) and support vector machines (SVM) are researched and compared and the SVM classification system shows better performance in binary classification tasks.
A joint spatial and spectral SVM&#x2019;s classification of panchromatic images
8 Citations2007Mathieu Fauvel, Jocelyn Chanussot +1 more
This paper investigates the use of an area filter to extract information about the inter-pixel dependency in the classification of very high resolution panchromatic images from urban areas using a support vector machines (SVM) classifier.