Standard Article

Theory and Algorithms of Perceptual Grouping

  1. Andrea Salgian

Published Online: 16 MAR 2009

DOI: 10.1002/9780470050118.ecse610

Wiley Encyclopedia of Computer Science and Engineering

Wiley Encyclopedia of Computer Science and Engineering

How to Cite

Salgian, A. 2009. Theory and Algorithms of Perceptual Grouping. Wiley Encyclopedia of Computer Science and Engineering. 2869–2875.

Author Information

  1. The College of New Jersey, Ewing, New Jersey

Publication History

  1. Published Online: 16 MAR 2009

Abstract

Perceptual grouping is the basic capability of the human brain to derive relevant structure without any prior content knowledge. This phenomenon has been described in the human visual system by Gestalt psychologists, and later, it was recognized by the computer vision community as a method for reducing the computational complexity of visual recognition. The article starts by introducing the laws of Gestalt, and it continues by describing how they are used to obtain perceptual grouping in computer vision. Different algorithms are presented for achieving perceptual grouping in two and three dimensions, at the signal, primitive, structural, and assembly levels. Spatiotemporal grouping is also discussed. The most successful computation techniques are described. Finally, a few methods for performance evaluation are also presented.

Keywords:

  • computer vision;
  • recognition;
  • perceptual grouping;
  • Gestalt laws;
  • 2D grouping;
  • 3D grouping;
  • spatiotemporal grouping;
  • performance evaluation