Abstract
Snapshot Compressive Imaging is an emerging technology that is based on compressive sensing theory to achieve high-efficiency hyperspectral data acquisition. The core problem of this technology is how to reconstruct 3D hyperspectral data from the 2D snapshot measurement in a fast and high-quality manner. In this paper, we propose a novel deep network, which consists of the symmetric residual module and the non-local spatial-spectral attention module, to learn the reconstruction mapping in a data-driven way. The symmetric residual module uses symmetric residual connections to improve the potential of interaction between convolution operations and further promotes the fusion of local features. The non-local spatial-spectral attention module is designed to capture the non-local spatial-spectral correlation in the hyperspectral image. Specifically, this module calculates the channel attention matrix to capture the global correlations between all of the spectral channels, and it fuses the channel attention attained feature maps and the spatial attention weighted features as the module output, thus both of the spatial-spectral correlations of hyperspectral images can be fully utilized for reconstruction. In addition, a compound loss, including the reconstruction loss, the measurement loss, and the cosine loss, is designed to guide the end-to-end network learning. We experimentally evaluate the proposed method on simulation and real datasets. The experimental results show that the proposed network outperforms the competing methods in terms of the reconstruction quality and running time.
Highlights
In this Simulation case, we use the same coded masks as in [26], which are from the real Coded Aperture Snapshot Spectral Imaging (CASSI) system and used to generate snapshot measurements of the test Hyperspectral images (HSIs)
The real data used in our experiments is the Bird data captured by the hyperspectral imaging camera
We propose a novel network for HSI snapshot reconstruction from a single measurement
Summary
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. Some commonly used priors include the sparse representation, the total variation (TV) [12], non-local similarity [13], and so on [14] Solving these problems requires the use of time-consuming iterative optimization, which leads to high reconstruction complexity. With the excellent learning ability of deep networks [15], scholars are committed to using deep convolutional networks to supervisely learn the explicit mapping from snapshot measurement to the original HSI This end-to-end learning method can significantly reduce the reconstruction time. In terms of spatial dimension, neighboring pixels usually have similar spectral characteristics For this reason, these prior structures should be used in the design of the network architecture, which can further improve the quality of reconstruction. The cosine loss can further enhance the fidelity of the reconstructed spectral signatures; 3. and experimental results demonstrate that the proposed model achieves better performance on simulation and real datasets, which proves the effectiveness and superiority of the proposed network
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
