Some Experimental Results on Linear Estimation for Image Analysis.

Abstract

Based on a constructed textured subimage, an experimental comparison is made on the three linear estimation methods for image segmentation, namely the maximum likelihood estimation, the maximum a posteriori estimation and the Fisher's linear discriminant. The last two methods are shown to be feasible for effective image segmentation. (Author)

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jan 28, 1981
Accession Number
ADA094281

Entities

People

  • Chia‐Hung Chen
  • Chihsung Yen
  • Rong-hwang Wu

Organizations

  • University of Massachusetts Dartmouth

Tags

Communities of Interest

  • Materials and Manufacturing Processes

DTIC Thesaurus Topics

  • Computations
  • Computer Vision
  • Databases
  • Electrical Engineering
  • Engineering
  • Feature Selection
  • Histograms
  • Image Segmentation
  • Iterations
  • Linear Filtering
  • Markov Chains
  • Maximum Likelihood Estimation
  • Military Research
  • Probability
  • Supervised Machine Learning
  • Two Dimensional

Readers

  • Computer Vision.
  • Regression Analysis.