Disparity Analysis of Time-Varying Imagery.

Abstract

An interactive program (called program DIDA) was developed to perform disparity analysis of time-varying imagery. The program consists of four sections: (1) image operations; (2) interesting point selection; (3) interesting point matching, and (4) image display. Three data bases were obtained: (1) a 20-frame sequence of images from a terrain board of the U.S. Army Night Vision Laboratory; (2) and 18-frame sequence of various moving objects from the University of Minnesota; and (3) a 33-frame sequence of images of a traffic scene from the University of Hamburg. Each of the four program sections were tested. Image operations included printing image intensities, and image file creation and deletion. Three interest operators were employed in the interesting point selection section: (1) simple variance; (2) directed variance (the Moravec Operator); and (3) edged variance. The number of interesting points selected was found to be inversely proportional to the interest operator threshold. A relaxation-labeling matching algorithm was used in the interesting point matching section to match interesting points from two images. Very poor matching results were obtained. Image displays included black-and-white and color images with and without the interesting points overwritten. Photographs were taken of both cases. (Author)

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Dec 01, 1981
Accession Number
ADA127409

Entities

People

  • Franklin D. Cooper Iii

Organizations

  • Air Force Institute of Technology

Tags

Communities of Interest

  • C4I
  • Sensors
  • Weapons Technologies

DTIC Thesaurus Topics

  • Air Force
  • Algorithms
  • Artificial Intelligence
  • Computer Graphics
  • Computer Programming
  • Computer Programs
  • Computer Science
  • Computer Vision
  • Computers
  • Databases
  • Detection
  • Image Processing
  • Images
  • Information Processing
  • Operating Systems
  • Pattern Recognition
  • Procedures (Computers)

Readers

  • Computer Science/Computer Engineering/Data Science/Digital Signal Processing.
  • Computer Vision.