Image Processing Applications for Nonlinear Decomposition and Synthesis

Abstract

Data representation, which is often overlooked in many image processing and analysis applications, is as critical as the algorithms applied to that data. When data is represented properly, simple algorithms can be much more powerful than sophisticated and often complex algorithms applied to an improper representation. In an image, the useful information is generally mixed in with irrelevant information or noise and often it is difficult for the computer to separate the useful information from the large volume of irrelevant data without destroying much of the useful data. To provide a solid foundation for a good solution to this problem, multiresolution decomposition and synthesis approaches have been developed which decompose raw image data into a set of partial information channels, where each channel represents a certain modality (or aspect) of the raw image. The channels can then be processed individually or cooperatively with a wide variety of results possible.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Aug 25, 1991
Accession Number
ADA243357

Entities

People

  • Eric W. Kelm

Organizations

  • Air Force Institute of Technology

Tags

Communities of Interest

  • Energy and Power Technologies
  • Materials and Manufacturing Processes

DTIC Thesaurus Topics

  • Artifacts
  • Compression
  • Compressive Properties
  • Data Compression
  • Digital Image Processing
  • Digital Images
  • Electrical Engineering
  • Frequency
  • Frequency Domain
  • Gray Scale
  • Image Compression
  • Image Processing
  • Image Reconstruction
  • Signal Processing
  • Three Dimensional
  • Two Dimensional
  • Visual Inspection

Readers

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