Skip to Main content Skip to Navigation
Conference papers

Detection of nonlinear mixtures using Gaussian processes: Application to hyperspectral imaging

Abstract : This paper investigates the use of Gaussian processes to detect non-linearly mixed pixels in hyperspectral images. The proposed technique is independent of nonlinear mixing mechanism, and therefore is not restricted to any prescribed nonlinear mixing model. The observed reflectances are estimated using both the least squares method and a Gaussian process. The fitting errors of the two approaches are combined in a test statistics for which it is possible to estimate a detection threshold given a required probability of false alarm. The proposed detector is compared to a robust nonlinearity detector recently proposed using synthetic data and is shown to provide a better detection performance. The new detector is also tested on a real hyperspectral image.
Complete list of metadatas

https://hal.archives-ouvertes.fr/hal-01484999
Contributor : Open Archive Toulouse Archive Ouverte (oatao) <>
Submitted on : Wednesday, March 8, 2017 - 9:55:44 AM
Last modification on : Wednesday, September 30, 2020 - 10:04:02 AM
Long-term archiving on: : Friday, June 9, 2017 - 12:52:09 PM

File

imbiriba_17115.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : hal-01484999, version 1
  • OATAO : 17115

Citation

Tales Imbiriba, José Carlos Bermudez, Jean-Yves Tourneret, Cédric Richard. Detection of nonlinear mixtures using Gaussian processes: Application to hyperspectral imaging. IEEE International Conference on Acoustics, Speech, and Signal Processing (ICASSP 2014), May 2014, Florence, Italy. pp. 7949-7953. ⟨hal-01484999⟩

Share

Metrics

Record views

261

Files downloads

329