Photon-Limited Image Detection Using Shot-Noise Models.

Abstract

The objective of this research continues to be the study of shot noise models and their application to the development of computationally feasible procedures for image detection problems. Also of interest is the performance evaluation of these procedures. These efforts are motivated by applications to low light level imaging as would occur in low dose x ray exposures or in night vision systems. Results are also be applicable to photon limited optical communication systems and to optical neural networks. Considerable progress has been made in understanding shot noise. The three major results of our research are (i) methods for computing shot noise distributions when both the intensity function and the system impulse response are known; (ii) methods for estimating an unknown intensity when only the impulse response is known; and (iii) methods for jointly estimating the intensity and the impulse response when both are unknown (a type of blind deconvolution).

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Aug 31, 1996
Accession Number
ADA319827

Entities

People

  • J. A. Gubner

Organizations

  • University of Connecticut

Tags

Communities of Interest

  • Human Systems

DTIC Thesaurus Topics

  • Avalanche Photodiodes
  • Communication Systems
  • Data Rate
  • Detection
  • Engineering
  • Information Theory
  • Low Light Levels
  • Multiple Access
  • Neural Networks
  • Night Vision
  • Optical Communications
  • Photons
  • Probability
  • Probability Distributions
  • Random Variables
  • Shot Noise
  • Signal Processing

Fields of Study

  • Physics

Readers

  • Explosive Engineering.
  • Image Processing and Computer Vision.
  • Optical Physics and Photonics.

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference