CODING COLOR PICTURES.

Abstract

A computer-simulation study of efficient coding for color pictures has been undertaken. Two typical color transparencies were resolved into three primaries, sampled in a square array and recored digitally on magnetic tape. The computer program transformed these data into luminance and chrominance quantities, performed certain parameter modifications, reconverted them into primary-color quantities and wrote them on an output tape. The parameters that were modified were the effective number of samples per picture and the number of quantum values each for the luminance and for the chrominance. The output tape was played back through the recorder-reproducer to produce images of the coded pictures on the face of the cathode-ray tube, which were photographed through appropriate filters on color film. The resultant transparencies were later viewed and compared by a number of observers to determine the absolute and relative quality achieved with the various codes (as affected by the variously modified parameters). Also, a test was run with a large number of observers to determine the relative recognizability of objects in one of the pictures when variously coded in color or monochromatically. Two major conclusions were drawn from this study. (i) A normal monochromatic picture can be converted into a full color picture of the same apparent sharpness by transmitting additionally only a fraction of a bit per sample. (Author)

Document Details

Document Type
Technical Report
Publication Date
Jun 17, 1964
Accession Number
AD0601666

Entities

People

  • Uri F. Gronemann

Organizations

  • Massachusetts Institute of Technology

Tags

DTIC Thesaurus Topics

  • Cathode Ray Tubes
  • Color Film
  • Computer Programming
  • Computer Programs
  • Computer Simulations
  • Computers
  • Luminance
  • Magnetic Tape
  • Observers
  • Photographic Materials
  • Photography
  • Recording Systems
  • Simulations
  • Simulators
  • Tapes
  • Transparencies

Fields of Study

  • Physics

Readers

  • Computer Science/Computer Engineering/Data Science/Digital Signal Processing.
  • Computer Vision.
  • Human-Computer Interaction (HCI).

Technology Areas

  • Quantum Computing