A comparative study of various meta-heuristic techniques applied to the multilevel thresholding problem

作者：Hammouche Kamal; Diaf Moussa; Siarry Patrick^*

来源：Engineering Applications of Artificial Intelligence, 2010, 23(5): 676-688.

DOI：10.1016/j.engappai.2009.09.011

• Multilevel thresholding
• Image segmentation
• Genetic algorithm
• Particle swarm optimization
• Differential evolution
• Ant colony optimization
• Simulated annealing
• Tabu search

The multilevel thresholding problem is often treated as a problem of optimization of an objective function. This paper presents both adaptation and comparison of six meta-heuristic techniques to solve the multilevel thresholding problem: a genetic algorithm, particle swarm optimization, differential evolution, ant colony, simulated annealing and tabu search. Experiments results show that the genetic algorithm, the particle swarm optimization and the differential evolution are much better in terms of precision, robustness and time convergence than the ant colony, simulated annealing and tabu search. Among the first three algorithms, the differential evolution is the most efficient with respect to the quality of the solution and the particle swarm optimization converges the most quickly.

• 出版日期2010-8