Recent Advances in Compressed Sensing: Discrete Uncertainty Principles and Fast Hyperspectral Imaging

Abstract

Compressed sensing is an important field with continuing advances in theory and applications. This thesis provides contributions to both theory and application. Much of the theory behind compressed sensing is based on uncertainty principles, which state that a signal cannot be concentrated in both time and frequency. We develop a new discrete uncertainty principle and use it to demonstrate a fundamental limitation of the demixing problem, and to provide a fast method of detecting sparse signals. The second half of this thesis focuses on a specific application of compressed sensing: hyperspectral imaging. Conventional hyperspectral platforms require long exposure times, which can limit their utility, and so we propose a compressed sensing platform to quickly sample hyperspectral data. We leverage certain combinatorial designs to build good coded apertures, and then we apply block orthogonal matching pursuit to quickly reconstruct the desired imagery.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Mar 26, 2015
Accession Number
ADA614340

Entities

People

  • Megan E. Lewis

Organizations

  • Air Force Institute of Technology

Tags

Communities of Interest

  • Biomedical
  • Energy and Power Technologies

DTIC Thesaurus Topics

  • Air Force
  • Algorithms
  • Charge Coupled Devices
  • Compressed Sensing
  • Detection
  • Detectors
  • Frequency
  • Hyperspectral Imagery
  • Integrals
  • Materials
  • Mathematics
  • Measurement
  • Spectra
  • Spectroscopy
  • Two Dimensional
  • United States
  • United States Government

Readers

  • Image Processing and Computer Vision.
  • Team-Based Human-Centered Cognitive Task Decision Making and Information Performance.