Improved genetic algorithm based 2-D maximum entropy image segmentation algorithm

Abstract

Since simple genetic algorithm based 2-D maximum entropy image segmentation algorithm has the problem of premature, this paper proposes an improved genetic algorithm. Through using Fitness Extreme Distance (FED), the improved genetic algorithm proposed in this paper establishes fuzzy evaluation mechanism in the evolution procedure. Compared with the simple genetic algorithm, improved algorithm remarkably enhances the genetic algorithm's convergence and the overall search ability. Theoretical analysis and experiment results shows, compared with basic genetic algorithm, the proposed genetic algorithm based 2-D Maximum Entropy Segmentation algorithm's acquired segmentation effectiveness is better. © 2012 Springer-Verlag GmbH.