Compressive Hyperspectral Imaging with Side Information

Abstract

A blind compressive sensing algorithm is proposed to reconstruct hyperspectral images from spectrally-compressed measurements.The wavelength-dependent data are coded and then superposed, mapping the three-dimensional hyperspectral datacube to a two-dimensional image. The inversion algorithm learns a dictionary {\em in situ} from the measurements via global-local shrinkage priors. By using RGB images as side information of the compressive sensing system, the proposed approach is extended to learn a coupled dictionary from the joint dataset of the compressed measurements and the corresponding RGB images, to improve reconstruction quality. A prototype camera is built using a liquid-crystal-on-silicon modulator. Experimental reconstructions of hyperspectral datacubes from both simulated and real compressed measurements demonstrate the efficacy of the proposed inversion algorithm, the feasibility of the camera and the benefit of side information.

20 pages, 21 figures. To appear in the IEEE Journal of Selected Topics Signal Processing

References in corpus (1)

• Classification and Reconstruction of High-Dimensional Signals from Low-Dimensional Features in the Presence of Side Information

Cited by in corpus (32)

• Rank Minimization for Snapshot Compressive Imaging
• Compressive Hyperspectral Imaging via Approximate Message Passing
• GAP-net for Snapshot Compressive Imaging
• Computational Imaging and Artificial Intelligence: The Next Revolution of Mobile Vision
• Generalized Alternating Projection Based Total Variation Minimization for Compressive Sensing
• Spectral ghost imaging camera with super-Rayleigh modulator
• Compressive spectral imaging based on hexagonal blue noise coded apertures
• Plug-and-Play Algorithms for Large-scale Snapshot Compressive Imaging
• Mask-guided Spectral-wise Transformer for Efficient Hyperspectral Image Reconstruction
• Key frames assisted hybrid encoding for photorealistic compressive video sensing
• Compressive Sensing via Low-Rank Gaussian Mixture Models
• A Dictionary-Based Generalization of Robust PCA Part II: Applications to Hyperspectral Demixing
• A Dictionary-Based Generalization of Robust PCA with Applications to Target Localization in Hyperspectral Imaging
• Lensless Compressive Imaging
• Image Compression Based on Compressive Sensing: End-to-End Comparison with JPEG
• Memory-Efficient Network for Large-scale Video Compressive Sensing
• Enhancement of CASSI by a zero-order image employing a single detector
• Nonlocal Low-Rank Tensor Factor Analysis for Image Restoration
• Deep Gaussian Scale Mixture Prior for Spectral Compressive Imaging
• Plug-and-Play Algorithms for Video Snapshot Compressive Imaging
• Fast Hyperspectral Image Recovery via Non-iterative Fusion of Dual-Camera Compressive Hyperspectral Imaging
• Self-supervised Neural Networks for Spectral Snapshot Compressive Imaging
• Various Total Variation for Snapshot Video Compressive Imaging
• Differential Coded Aperture Single-Snapshot Spectral Imaging
• Unsupervised Spatial-spectral Network Learning for Hyperspectral Compressive Snapshot Reconstruction
• MetaSCI: Scalable and Adaptive Reconstruction for Video Compressive Sensing
• Deep-learning-based Hyperspectral imaging through a RGB camera
• Compressive Sensing via Convolutional Factor Analysis
• Approximate Message Passing in Coded Aperture Snapshot Spectral Imaging
• Convergence of the Generalized Alternating Projection Algorithm for Compressive Sensing
• Snapshot compressed sensing: performance bounds and algorithms
• Deep Unfolding-Aided Parameter Tuning for Plug-and-Play-Based Video Snapshot Compressive Imaging