摘要
为改善传统图像融合方法对细节信息的丢失,提出了一种基于遗传粒子群算法(genetic algorithm of particleswarm optimization,GAPSO)的图像融合方法,该算法应用于像素级的图像融合,使图像融合问题归结为最优化问题.该算法结合遗传算法和粒子群算法的优点,对标准粒子群算法进行了改进,将交叉与变异算子引入到标准粒子群算法,提高了该算法的收敛性能和全局求解能力.实验结果表明,该算法获得的评价指标都优于遗传算法和PSO算法,且融合图像较好地从源图像中提取了有用信息,提高了融合质量.
To avoid the loss of detailed information in processes of traditional image fusion, based on GAPSO（genetic algorithm of particle swarm optimization）, a new image fusion approach is proposed, which brings the image fusion into an optimization problem. Our algorithm embodies the advantages of both genetic and particle swarm algorithms. The cross operator and mutation operator are introduced into every iteration, which improves the convergency and the global ability of PSO（particle swarm optimization） algorithm. The experimental results show that the proposed method can extract the useful information from original images well and enhance the fusion quality, all evaluating in- dices are better than those of genetic algorithm and PSO algorithm.
出处
《徐州师范大学学报（自然科学版）》
CAS
2010年第1期55-58,共4页
Journal of Xuzhou Normal University(Natural Science Edition)
关键词
遗传粒子群算法
像素
图像融合
genetic algorithm of particle swarm optimization（GAPSO）
pixel
image fusion
作者简介
作者简介：彭圣华，女，讲师．
