Model-Based Building Verification in Aerial Photographs.

Abstract

This paper concerns the design of a computer vision system for change detection. Here, change detection is defined as figuring out the differences between an object model and the newly sensed image. The target objects are confined to the cultural features, such as roads and buildings. We divide the task into two modules: model verification and image interpretation. In this report, the verification stage will be discussed in detail. In general there exists a lot of domain specific heuristics to judge the status of changes. For example, to verify the existence of a building, we can check its shape, size, height, surface direction, and surface material, etc. The expert system approach is a natural approach which can code all information together. while different photo interpreters and field specialists may have different viewpoints about the status of an object, expert systems can be modified easily to reflect a particular one's viewpoint.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Sep 01, 1987
Accession Number
ADA186010

Entities

People

  • Chin-hwa Lee
  • Hsi-jian Lee

Organizations

  • Naval Postgraduate School

Tags

Communities of Interest

  • Air Platforms
  • Energy and Power Technologies

DTIC Thesaurus Topics

  • Aerial Photographs
  • Artificial Intelligence
  • Change Detection
  • Computer Science
  • Computer Vision
  • Computers
  • Control Systems
  • Coordinate Systems
  • Detection
  • Electrical Engineering
  • Engineering
  • Expert Systems
  • Image Processing
  • Image Segmentation
  • Language
  • Three Dimensional
  • Two Dimensional

Readers

  • Computer Vision.
  • Software Engineering.

Technology Areas

  • AI & ML