Kernel Principal Component Analysis for the Classification of Hyperspectral Remote Sensing Data over Urban Areas

By Mathieu Fauvel, Jocelyn Chanussot, Jon Atli Benediktsson. EURASIP Journal on Advances in Signal Processing, 2009

296 citations recorded by the source corpus.

Kernel principal component analysis (KPCA) is investigated for feature extraction from hyperspectral remote sensing data. Features extracted using KPCA are classified using linear support vector machines. In one experiment, it is shown that kernel principal component features are more linearly separable than features extracted with conventional principal component analysis. In a second experiment, kernel principal components are used to construct the extended morphological profile (EMP). Classification results, in terms of accuracy, are improved in comparison to original approach which used conventional principal component analysis for constructing the EMP. Experimental results presented in this paper confirm the usefulness of the KPCA for the analysis of hyperspectral data. For the one data set, the overall classification accuracy increases from 79% to 96% with the proposed approach.

Subfields: Analytical Chemistry, Atmospheric Science, Media Technology

View Paper: https://www.semanticscholar.org/paper/7e0545c0d6613d7df4493f96bc3745243c9c4d39

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Analytical Chemistry

In Atmospheric Science

In Media Technology

Citations in the corpus

Feature Extraction for Hyperspectral Imagery: The Evolution From Shallow to Deep: Overview and Toolbox (2020, methodology): “KICA and KPCA have been used as UFE techniques for change detection and classification in [28] and [29], respectively.”
A Survey on Spectral–Spatial Classification Techniques Based on Attribute Profiles (2015, background): “The second issue can be classically overcome by considering cross-validation techniques using training data [67].”
Collaborative Representation for Hyperspectral Anomaly Detection (2015, background): “In mathematical modeling, the observation data X can be expressed as X = L+ S (15) 1http://perception.csl.illinois.edu/matrix-rank/sample_code.html. where L is a low-rank matrix and S is a sparse matrix of entries with a small fraction of nonzero entries.”
Support vector machines in remote sensing: A review (2011, unknown): “…is assumed to match the number of spectral bands, were normally distributed, the assumption that the distribution can be described using a bell-shaped (Gaussian) function ceases to be sound, since the concentration of data in higher dimensional space tends to be in the tails (Fauvel et al., 2009).”
A Novel Ranking-Based Clustering Approach for Hyperspectral Band Selection (2016, methodology): “as projection pursuit [13], [14], principal component analysis (PCA) [15]–[17], and independent component analysis [18], whereas the latter identifies a subset of the original spectral bands that contains most of the characteristics [19].”
Linear Versus Nonlinear PCA for the Classification of Hyperspectral Data Based on the Extended Morphological Profiles (2012, background): “(KPCA) instead of the features from PCA [8].”
Generalized Composite Kernel Framework for Hyperspectral Image Classification (2013, background): “The good generalization capability of machine learning techniques such as the support vector machine (SVM) [6] and multinomial logistic regression (MLR) [8]–[12] can still be enhanced by reducing the input data dimensionality [13], [14], which can help in addressing ill-posed problems based on limit”