Hyperspectral Compressive Sensing Imaging via Spectral Sparse Constraint

Qi Wang, Hong Xu, Lingling Ma, Chuanrong Li, Yongsheng Zhou, Lingli Tang

2018

Abstract

The existing algorithms to reconstruct hyperspectral compressive sensing images mainly use the sparse property of spatial information and some simple non-adaptive spectral constraint such as the low-rank property. However, these strategies cannot remove the spectral redundancy efficiently and a new method to make full use of the abundant redundancy of spectral information and improve the quality for hyperspectral CS reconstruction is necessary. A new CS sampling and reconstruction model based on spectral sparse representation was proposed in this paper. The spectral sparse dictionary was constructed from training samples to enhance the effect of sparse representation and the total variation constraint of spatial images was also considered to further enhance the precision during the reconstruction. The experiment to reconstruct AVIRIS hyperspectral images of 200 bands show that the hyperspectral image was almost perfectly reconstructed at 25% sampling rate and the spatial and spectral precision was higher than traditional methods which only adopt the spatial sparsity and simple non-adaptive spectral constraint in the same condition.

Download


Paper Citation


in Harvard Style

Wang Q., Xu H., Ma L., Li C., Zhou Y. and Tang L. (2018). Hyperspectral Compressive Sensing Imaging via Spectral Sparse Constraint.In Proceedings of the 6th International Conference on Photonics, Optics and Laser Technology - Volume 1: PHOTOPTICS, ISBN 978-989-758-286-8, pages 273-278. DOI: 10.5220/0006664202730278


in Bibtex Style

@conference{photoptics18,
author={Qi Wang and Hong Xu and Lingling Ma and Chuanrong Li and Yongsheng Zhou and Lingli Tang},
title={Hyperspectral Compressive Sensing Imaging via Spectral Sparse Constraint},
booktitle={Proceedings of the 6th International Conference on Photonics, Optics and Laser Technology - Volume 1: PHOTOPTICS,},
year={2018},
pages={273-278},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0006664202730278},
isbn={978-989-758-286-8},
}


in EndNote Style

TY - CONF

JO - Proceedings of the 6th International Conference on Photonics, Optics and Laser Technology - Volume 1: PHOTOPTICS,
TI - Hyperspectral Compressive Sensing Imaging via Spectral Sparse Constraint
SN - 978-989-758-286-8
AU - Wang Q.
AU - Xu H.
AU - Ma L.
AU - Li C.
AU - Zhou Y.
AU - Tang L.
PY - 2018
SP - 273
EP - 278
DO - 10.5220/0006664202730278