KRISM—Krylov Subspace-based Optical Computing of Hyperspectral Images

@article{Saragadam2019KRISMKrylovSO,
  title={KRISM—Krylov Subspace-based Optical Computing of Hyperspectral Images},
  author={Vishwanath Saragadam and Aswin C. Sankaranarayanan},
  journal={ACM Transactions on Graphics (TOG)},
  year={2019},
  volume={38},
  pages={1 - 14}
}
We present an adaptive imaging technique that optically computes a low-rank approximation of a scene’s hyperspectral image, conceptualized as a matrix. Central to the proposed technique is the optical implementation of two measurement operators: a spectrally coded imager and a spatially coded spectrometer. By iterating between the two operators, we show that the top singular vectors and singular values of a hyperspectral image can be adaptively and optically computed with only a few iterations… Expand
SASSI — Super-Pixelated Adaptive Spatio-Spectral Imaging
TLDR
A scene-adaptive spatial sampling of a hyperspectral scene, guided by its super-pixel segmented image, is capable of obtaining high-quality reconstructions and validated with extensive simulations as well as a lab prototype. Expand
Hyperspectral Image Super-Resolution Based on Spatial Group Sparsity Regularization Unmixing
TLDR
This paper proposes a spatial group sparsity regularization unmixing-based method for hyperspectral super-resolution and demonstrates that the proposed method has superior performance when compared to the existing state-of-the-art. Expand
Programmable Spectrometry - Per-pixel Classification of Materials using Learned Spectral Filters
TLDR
This work proposes a novel programmable camera that is capable of producing images of a scene with an arbitrary spectral filter and uses this camera to optically implement the spectral filtering of the scene's hyperspectral image with the bank of spectral profiles needed to perform per-pixel material classification. Expand
An overview of joint activities on computational imaging and compressive sensing systems by NATO SET-232
TLDR
An overview of the ongoing joint activities by NATO SET-232, current computational imaging and compressive sensing technologies, limitations of the design trade space, algorithm and conceptual design considerations, and field performance assessment and modeling is presented. Expand
Hyperspectral Image Compressive Sensing Reconstruction Using Subspace-Based Nonlocal Tensor Ring Decomposition
TLDR
A novel subspace-based nonlocal tensor ring decomposition method (SNLTR) for HSI-CSR that assumes that the HSI lies in a low-dimensional subspace to explore the nonlocal self-similarity and preserve the spatial structure of HSI. Expand
Spectrally-Programmable Cameras for Imaging and Inference
TLDR
The work in this thesis seeks to simplify the process of capturing spectral information of a scene by building novel and effi?cientoptical systems for imaging and inference, and lays groundwork for making hyperspectral cameras more practical by introducing computing into the sensing pipeline, and moving most of computational burden into the optical domain. Expand
On space-spectrum uncertainty analysis for spectrally programmable cameras.
TLDR
It is not possible to simultaneously acquire high-resolution spatial images while programming the spectrum at high resolution due to a Fourier relationship between the aperture used for resolving spectrum and its corresponding diffraction blur in the spatial image. Expand
Deconvolving Diffraction for Fast Imaging of Sparse Scenes
TLDR
The presented combination of the imaging approach and scene recovery method is well suited for high-speed marker based motion capture and particle image velocimetry (PIV) over long periods. Expand
Diffraction Line Imaging
TLDR
A novel computational imaging principle that combines diffractive optics with line (1D) sensing that is able to achieve high-speed and high-accuracy 2D positioning of light sources and 3D scanning of scenes is presented. Expand
Image-based acquisition and modeling of polarimetric reflectance
TLDR
This work provides the first polarimetric BRDF (pBRDF) dataset that captures the polarIMetric properties of real-world materials over the full angular domain, and at multiple wavelengths, and proposes a scheme combining image-based acquisition with spectroscopic ellipsometry to perform measurements in a realistic time. Expand
...
1
2
...

References

SHOWING 1-10 OF 104 REFERENCES
One-Shot Hyperspectral Imaging Using Faced Reflectors
TLDR
This paper proposes a novel technique of one-shot hyperspectral imaging using faced reflectors on which color filters are attached, based on the principle that each of multiple reflections on the filters has a different spectrum, which allows for multiple intensities through different spectra. Expand
Statistics of real-world hyperspectral images
TLDR
Using a new collection of fifty hyperspectral images of indoor and outdoor scenes, an optimized “spatio-spectral basis” is derived for representing hyperspectrals image patches and statistical models for the coefficients in this basis are explored. Expand
Dual-coded compressive hyperspectral imaging.
TLDR
It is demonstrated that two high-speed spatial light modulators, located conjugate to the image and spectral plane, respectively, can code the hyperspectral datacube into a single sensor image such that the high-resolution signal can be recovered in postprocessing. Expand
Compact single-shot hyperspectral imaging using a prism
TLDR
An image formation model that can predict the perspective projection of dispersion, and a reconstruction method that can estimate the full spectral information of a scene from sparse dispersion information are formulated. Expand
High-quality hyperspectral reconstruction using a spectral prior
TLDR
A novel hyperspectral image reconstruction algorithm, which overcomes the long-standing tradeoff between spectral accuracy and spatial resolution in existing compressive imaging approaches and introduces a novel optimization method, which jointly regularizes the fidelity of the learned nonlinear spectral representations and the sparsity of gradients in the spatial domain. Expand
Sparse Recovery of Hyperspectral Signal from Natural RGB Images
TLDR
This work presents a low cost and fast method to recover high quality hyperspectral images directly from RGB using a novel, larger-than-ever database of hyperspectRAL images serves as a hyperspectrals prior. Expand
Spatial-spectral encoded compressive hyperspectral imaging
TLDR
This paper proposes a novel compressive hyperspectral (HS) imaging approach that allows for high-resolution HS images to be captured in a single image and demonstrates other applications for the over-complete HS dictionary and sparse coding techniques, including 3D HS images compression and denoising. Expand
Hyperspectral image compressed sensing via low-rank and joint-sparse matrix recovery
TLDR
This work proposes a novel approach to reconstruct Hyperspectral images from very few number of noisy compressive measurements based on a convex minimization which penalizes both the nuclear norm and the ℓ2,1 mixed-norm of the data matrix. Expand
A Compressive Sensing and Unmixing Scheme for Hyperspectral Data Processing
TLDR
Experimental and computational evidences obtained from this paper indicate that the proposed scheme for hyperspectral data compression and reconstruction has a high potential in real-world applications. Expand
Hyperspectral image classification and dimensionality reduction: an orthogonal subspace projection approach
TLDR
A technique which simultaneously reduces the data dimensionality, suppresses undesired or interfering spectral signatures, and detects the presence of a spectral signature of interest is described. Expand
...
1
2
3
4
5
...