An Examination of Several Methods of Hyperspectral Image Denoising: Over Channels, Spectral Functions and Both Domains
The corruption of hyperspectral images by noise can compromise tasks such as classification, target detection and material mapping. For this reason, many methods have been proposed to recover, as best as possible, the uncorrupted hyperspectral data from a given noisy observation. In this paper, we propose and compare the results of four denoising methods which differ in the way the hyperspectral data is treated: (i) as 3D data sets, (ii) as collections of frequency bands and (iii) as collections of spectral functions. In the case of additive noise, these methods can be easily adapted to accommodate different degradation models. Our methods and results help to address the question of how hyperspectral data sets should be processed in order to obtain useful denoising results.
KeywordsSpectral Function Sparse Representation Hyperspectral Image Spectral Domain Hyperspectral Data
Unable to display preview. Download preview PDF.
- 1.AVIRIS website: http://aviris.jpl.nasa.gov/index.html
- 12.Donoho, D., Johnstone, I.M., Kerkyacharian, G., Picard, D.: Wavelet shrinkage: asymptopia. Journal of the Royal Statistical Society, Ser. B, 371–394 (1995)Google Scholar
- 17.Mairal, J., Bach, F., Jenatton, R., Obozinski, G.: Convex Optimization with Sparsity-Inducing Norms. Optimization for Machine Learning. MIT Press (2011)Google Scholar
- 21.Rasti, B., Sveinsson, J.R., Ulfarsson, M.O., Benediktsson, J.A.: Hyperspectral image denoising using 3D wavelets. In: 2012 IEEE International Geoscience and Remote Sensing Symposium (IGARSS), pp. 1349–1352 (2012)Google Scholar
- 23.Shippert, P.: Why use Hyperspectral imaging? Photogrammetric Engineering and Remote Sensing, 377–380 (2004)Google Scholar
- 24.Turlach, B.A.: On algorithms for solving least squares problems under an \(\ell _1\) penalty or an \(\ell _1\) constraint. In: Proceedings of the American Statistical Association, Statistical Computing Section, pp. 2572–2577 (2005)Google Scholar