Computer Engineering and Applications ›› 2009, Vol. 45 ›› Issue (3): 175-177.DOI: 10.3778/j.issn.1002-8331.2009.03.052

• 图形、图像、模式识别 • Previous Articles     Next Articles

Improved Particle Swarm Optimization with dynamically changing inertia weight

FENG Ting,LU Xue-song,YANG Wei,ZHANG Su   

  1. Biomedical Instrument Institute,Shanghai Jiaotong University,Shanghai 200240,China
  • Received:2008-07-21 Revised:2008-09-24 Online:2009-01-21 Published:2009-01-21
  • Contact: FENG Ting

改进收敛条件的动态调整惯性权重PSO算法

冯 婷,陆雪松,阳 维,张 素   

  1. 上海交通大学 生命科学技术学院 生物医学仪器研究所,上海 200240
  • 通讯作者: 冯 婷

Abstract: Particle Swarm Optimization (PSO) is a global evolutionary approach,which can effectively avoid the local extremum in biomedical image registration.A new particle swarm optimization with dynamically changing inertia weight is applied in image registration based on mutual information.The initial location is assigned uniformly to avoid local extremum caused by evolution in some little area initiated randomly.A parameter is added to control the ceasing of iteration,improving the speed.Experimental results indicate that the new algorithm can find the global optimization and converged fast.

Key words: image registration, Particle Swarm Optimization(PSO), dynamically changing inertia weight

摘要: 在医学图像配准中需要解决互信息图像配准过程中局部极值问题,引入了一种动态调整惯性权的自适应粒子群算法;验证了其中两个重要参数的取值,并均匀赋值粒子初始位置,避免随机产生的初始位置集中在某一区域而使寻优陷入局部极值,同时加入进化速度因子作为搜索中止条件,加快了搜索速度。实验表明,该算法既能找到全局最优又能快速收敛。

关键词: 医学图像配准, 全局寻优, 粒子群算法, 动态调整惯性权重