On a Two-Dimensional Maximum Entropy Spectral Estimation Method for the Texture-Image Analysis.

Abstract

The two-dimensional maximum entropy spectral estimation method proposed by Lim and Malik is studied and modified for spectral analysis of texture images. The computational efficiency of the algorithm makes it feasible for minicomputer implementation to extract effective spectral domain texture features for texture discrimination. Extensive computer results are presented.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Oct 30, 1981
Accession Number
ADA107569

Entities

People

  • Chia‐Hung Chen
  • Gia-kinh Young

Organizations

  • University of Massachusetts Dartmouth

Tags

Communities of Interest

  • Energy and Power Technologies

DTIC Thesaurus Topics

  • Algorithms
  • Autocorrelation
  • Coefficients
  • Computational Complexity
  • Databases
  • Discrimination
  • Electrical Engineering
  • Engineering
  • Equations
  • Human-Machine Interaction
  • Midrange Computers
  • Power Series
  • Power Spectra
  • Signal Processing
  • Spectra
  • Three Dimensional
  • Two Dimensional

Readers

  • Computer Vision.
  • Statistical inference.