Paper 14:

Genetic algorithms in image processing - A review

Catherine Bounsaythip Jarmo Alander
Centre d'Automatique de Lille University of Vaasa
U.S.T.L. (Lille1), Bat. P2 P.O. Box 700
59655 Villeneuve d'Ascq - France 65101 Vaasa - Finland
E-mail: [email protected] E-mail: [email protected]

Abstract. This paper reviews some applications of Genetic Algorithms (GAs) which have been adopted in the domain of image processing. Genetic Algorithms or generally Evolutionary Algorithms (EA) have been successfully applied to various oelds of optics and image processing (see the indexed bibliography by J. T. Alander [14]).

Image processing requires robust and fast techniques capable of managing large, noisy and fuzzy quantities of data. Evolutionary paradigms seem well suited to such kind of problems, in the sense that they usually do not need a priori domain specioc information.

This paper gives a review of the implementation of EA techniques ranging from low-level image processing to synthetic image animation.

Keywords: genetic algorithms, animation, classiocation, feature extraction, oltering /2D, image analysis, image processing, image segmentation, pattern recognition,

14.1 Introduction

Image processing and analysis have many important applications in an extremely diverse oeld of technologies including printing, TV, multimedia, aerial surveillance, satellite and space imaging, machine vision, security surveillance, quality control, medical image analysis, onger print analysis, character recognition etc.

One of the main objectives of image processing is to analyse images provided by an imaging system and then to locate and recognize the object in the environment. For example, navigation through an uncertain environment involves

173