An Image Processing Technique for Achieving Lossy Compression of Data at Ratios in Excess of 100:1

Abstract

This report presents a new method for image compression that achieves ratios in excess of 100: 1. It begins as a tutorial of basic compression techniques that includes both lossless and lossy methods. The term 'lossless' means that the compressed file can be reconstructed exactly, while lossy means that there is some loss of information in the reconstructed image. The most important components of the Joint Photographic Experts Group (JPEG) algorithm are targeted including Huffman encoding and the Cosine Transform. LZW, singular value decomposition, fractal compression, and other techniques are also discussed. The feasibility of using these methods for bandwidth reduction is analyzed relative to several High Definition TV (HDTV) standards. Finally, the cosine transform is modified to show how scanned images can be compressed 100: 1 with little loss in quality. Several examples are presented.... Compression, Lossy, Image processing, Video, Lossless.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Nov 01, 1992
Accession Number
ADA259218

Entities

People

  • Michael G. Ellis

Tags

Communities of Interest

  • Air Platforms
  • Energy and Power Technologies

DTIC Thesaurus Topics

  • Algorithms
  • Amplitude Modulation
  • Bandwidth
  • Coding
  • Compression
  • Computer Programming
  • Computer Programs
  • Computers
  • Data Compression
  • Decoding
  • Image Compression
  • Image Processing
  • Information Theory
  • Intellectual Property
  • Operating Systems
  • Standards
  • Two Dimensional

Readers

  • Image Processing and Computer Vision.
  • Systems Analysis and Design