Abstract—This paper presents the framework of kernel-based methods in the context of hyperspectral image classification, illustrating from a general viewpoint the main characteristics of different kernel-based approaches and analyzing their properties in the hyperspectral domain. In particular, we assess performance of regularized radial basis function neural networks (Reg-RBFNN), standard support vector machines (SVMs), kernel Fisher discriminant (KFD) analysis, and regularized AdaBoost (Reg-AB). The novelty of this work consists in: 1) introducing Reg-RBFNN and Reg-AB for hyperspectral image classification; 2) comparing kernel-based methods by taking into account the peculiarities of hyperspectral images; and 3) clarifying their theoretical relationships. To these purposes, we focus on the accuracy of methods when working in noisy environments, high input dimension, and limited training sets. In addition, some other important issues are discussed, such as the sparsity of the solutions, the computational burden, and the capability of the methods to provide outputs that can be directly interpreted as probabilities. Index Terms—AdaBoost, feature space, hyperspectral classification, kernel-based methods, kernel Fisher discriminant analysis, radial basis function neural networks, regularization, support vector machines. I.

Abstract — In this paper, we propose the use of Support Vector Machines (SVM) for automatic hyperspectral data classification and knowledge discovery. In the first stage of the study, we use SVMs for crop classification and analyze their performance in terms of efficiency and robustness, as compared to extensively used neural and fuzzy methods. Efficiency is assessed by evaluating accuracy and statistical differences in several scenes. Robustness is analyzed in terms of (a) suitability to working conditions when a feature selection stage is not possible, and (b) performance when different levels of Gaussian noise are introduced at their inputs. In the second stage of this work, we analyze the distribution of the support vectors (SV) and perform sensitivity analysis on the best classifier in order to analyze the significance of the input spectral bands. For classification purposes, six hyperspectral images acquired with the 128-band HyMAP spectrometer during the DAISEX-1999 campaign are used. Six crop classes were labelled for each image. A reduced set of labelled samples is used to train the models and the entire images are used to assess their performance. Several conclusions are drawn: (1) SVMs yield better outcomes than neural networks regarding accuracy, simplicity and robustness; (2) training neural and neurofuzzy models is unfeasible when working with high dimensional input spaces and great amounts of training data; (3) SVMs perform similarly for different training subsets with varying input dimension, which indicates that noisy bands are successfully detected; and (4) a valuable ranking of bands through sensitivity analysis is achieved. Index Terms — Hyperspectral imagery, crop classification, knowledge discovery, Support Vector Machines, neural networks.

Abstract. In this communication, we analyze several regularized types ofRadial Basis Function (RBF) Networks for crop classification using hyperspectral images. We compare the regularized RBF neural network with Support Vector Machines (SVM) using the RBF kernel, and AdaBoost Regularized (ABR) algorithm using RBF bases, in terms ofaccuracy and robustness. Several scenarios ofincreasing input space dimensionality (128, 6, 3 and 2 number ofbands) are tested for six labeled images containing six crop classes. Also, regularization, sparseness, and knowledge extraction are paid attention. Several conclusions are drawn: (1) all models offer similar accuracy but SVM and ABR yield slightly better results than RBFNN; (2) results indicate that ABR are less affected by the curse ofdimensionality and has identified efficiently the presence ofnoisy bands; (3) we find that regularization is a useful method to work with noisy data distributions; and (4) some physical consequences are extracted from the trained models. Finally, this preliminary work lead us to think ofkernel-based machines as efficient and robust methods for hyperspectral data classification. 1

Several modifications to parameter estimation in a Radial Basis Functions network are introduced. These include a better initializing clustering algorithm and a full gradient descent on centers and weights after weights were found via a matrix inversion. Performance comparison with other RBF algorithms is given on several data-sets. It is found that The proposed method was found superior to Bishop's EM training algorithm, to Orr's method [1] for as well as a conventional implementation. I. Introduction Radial basis functions have been extensively used for interpolation [2], [3], [4], [5], [6], [7] regression and classification due to their universal approximation properties and simple parameter estimation. The parameter estimation requires a (pseudo) inversion of a (possibly sparse) matrix. The possible numerical instability of the inversion (which is aggravated when the number of training patterns is small compared to the dimensionality) may be partially alleviated by further parame...

Abstract — In this study we apply a variant of a recently proposed linear subspace method, the Neighbourhood Component Analysis (NCA), to the task of hyperspectral classification. The NCA algorithm explicitly utilizes the classification performance criterion to obtain the optimal linear projection. NCA assumes nothing about the form of the each class and the shape of the separating surfaces. Experimental studies were conducted on the basis of hyperspectral images acquired by two sensors: the Airborne Visible/Infrared Imaging Spectroradiometer (AVIRIS) and AISA-EAGLE. Experimental results confirm the significant superiority of the NCA classifier in the context of hyperspectral data classification over methodologies that were previously suggested. Index Terms- Classification, hyperspectral images, remote sensing, linear projection, NCA. I.

Abstract not found