Modular Implementation of Feature Extraction and Matching Algorithms for Photogrammetric Stereo Imagery

Abstract

This paper describes the implementation of algorithms for automatically extracting and matching features in stereo pairs of images. The implementation has been designed to be as modular as possible to allow different algorithms for each stage in the matching process to be combined in the most appropriate manner for each particular problem. The modules have been implemented in the AVS environment but are designed to be portable to any platform. This work has been undertaken as part of task DEF 93/1 63 'Intelligence Analysis of Imagery', and forms part of ITD's contribution to the Visual Processing research program in the Centre for Sensor System and Information Processing. A major aim of both the task and the research program is to produce software to assist intelligence analysts in extracting three dimensional shape from imagery: the algorithms and software described here will form the first part of a module for automatically extracting depth information from stereo image pairs.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Jun 01, 1994
Accession Number
ADA285422

Entities

People

  • Garry Hamlyn
  • James Kershaw

Tags

Communities of Interest

  • C4I
  • Electronic Warfare
  • Energy and Power Technologies
  • Engineered Resilient Systems
  • Materials and Manufacturing Processes
  • Weapons Technologies

DTIC Thesaurus Topics

  • Acquisition
  • Computer Programs
  • Computer Vision
  • Computers
  • Feature Extraction
  • Flow Fields
  • Geometry
  • Gray Scale
  • Image Processing
  • Information Processing
  • Information Systems
  • Intelligence Analysis
  • Lists (Data Structures)
  • Operating Systems
  • Shape
  • Three Dimensional
  • Two Dimensional

Fields of Study

  • Computer science

Readers

  • Computer Vision.
  • Software Engineering

Technology Areas

  • AI & ML
  • AI & ML - Information Retrieval