On Computing Complete Histograms of Images in Log (n) Steps Using Hypercubes

  • Authors:
  • T. Bestul;L. S. Davis

  • Affiliations:
  • -;-

  • Venue:
  • IEEE Transactions on Pattern Analysis and Machine Intelligence
  • Year:
  • 1989

Quantified Score

Hi-index 0.14

Visualization

Abstract

An algorithm for the computation of the histogram of limited-width (such as gray-level) values on a SIMD (single-instructions multiple-data) hypercube multiprocessor is proposed, which does not require the use of a general interconnection capability such as that on the connection machine. The computation of the complete histogram of n such values takes place in a series of log n steps, after which the histogram for value i can be found in the lowest-addressed processor whose address ends in i. The algorithm makes use of the association of suffixes of data values of increasing width with suffixes of processor addresses.