[1]赵春晖,肖晓俊.克隆粒子群算法的镜像层叠滤波器的优化设计[J].智能系统学报,2008,3(2):129-134.
ZHAO Chun-hui,XIAO Xiao-jun.Optimizing stacked filters with mirrored threshold decomposition using an improved clone selection and PSO algorithm[J].CAAI Transactions on Intelligent Systems,2008,3(2):129-134.
点击复制
ZHAO Chun-hui,XIAO Xiao-jun.Optimizing stacked filters with mirrored threshold decomposition using an improved clone selection and PSO algorithm[J].CAAI Transactions on Intelligent Systems,2008,3(2):129-134.
克隆粒子群算法的镜像层叠滤波器的优化设计
《智能系统学报》[ISSN 1673-4785/CN 23-1538/TP] 卷:
3
期数:
2008年第2期
页码:
129-134
栏目:
学术论文—智能系统
出版日期:
2008-04-25
- Title:
- Optimizing stacked filters with mirrored threshold decomposition using an improved clone selection and PSO algorithm
- 文章编号:
- 1673-4785(2008)02-0129-06
- 关键词:
- 镜像阈值分解; 正布尔函数(PBF); 层叠滤波器; 克隆选择; 粒子群
- Keywords:
- mirrored threshold decomposition; positive Boolean function (PBF); stacked filters; clone selection algorithm; Particle swarm optimization(PSO)
- 分类号:
- TN911.26
- 文献标志码:
- A
- 摘要:
- 层叠滤波器是一种具有层叠性和阈值分解性的滑动窗非线性数字滤波器.层叠滤波器的优化设计过程关键就是对正布尔函数优化.本文利用了克隆粒子群优化算法的全局寻优能力和免疫系统的免疫信息处理机制，改善了粒子群优化算法摆脱局部极值点的能力，提高了算法进化过程中的收敛速度和精度.仿真实验的结果也表明,该算法在正布尔函数的优化效果得到了显著改善，收敛加快，试验结果表明该算法设计的层叠滤波器具有良好的细节保持能力和去噪声能力，有效地改善了滤波性能.
- Abstract:
- Stack filters are a class of nonlinear digital filters with sliding windows. Their two main properties are threshold decomposition and stacking. Optimal design of these filters requires optimization of a positive Boolean function (PBF). Optimization of stacked filters using traditional algorithms is easily trapped in a local optimum. To prevent this in recent years the bionic clone selection and PSO algorithms were proposed. The combination of these two algorithms avoids the problem of falling into a local minimum, greatly improving convergence speed and accuracy．Experimental results show that optimally stacked filters effectively suppress noise while retaining details of images and enhancing filtering capability.
- 参考文献/References:
-
［1］COYLE E J, LIN J H. Stack filters and the mean absolute error criter ion［J］. IEEE Transaction on Signal Processing, 1988, 36(8): 12441254.
［2］PARADES J L, ARCE G R. Optimization of stack filters based on mirrored thr eshold decomposition［J］. IEEE Transaction on Signal Procession, 2001, 49(6): 1 1791188.
［3］崔 颖, 赵春晖.基于镜像阈值分解的层叠滤波器优化［J］.哈尔滨工程大学学报, 2006, 27(6): 904907.
CUI Ying, ZHAO Chunhui. Optimization of stack filters based on mirrored thre shold decomposition［J］. Journal of Harbin Engineering University ,2006, 27(6): 904907.
［4］赵春晖,孙 莉, 付正威. 基于克隆选择算法的层叠滤波器的优化设计［J］. 哈尔滨工程大学学报, 2007, 28(4):454460.
ZHAO Chunhui, SUN Li, FU Zhengwei. Optimizing stack filters through a clone selection algorithm［J］. Journal of Harbin Engineering University, 2007, 28(4): 454460.
［5］PARADES J L, ARCE G R. Stack filters, stack smoothers, and mirrored threshold decomposition［J］. IEEE Transactions on Signal Processing, 1999, 47( 10): 27572767.
［6］吴静敏, 左洪福. 基于免疫粒子群算法的组合预测方法［J］.系统工程理论方法应用, 2006, 15(3): 229233.
WU Jingmin,ZUO Hongfu. A combined forecasting method base on PSO with immunity a lgorithms ［J］. Systems Engineering Theory ?Methodology Applications, 2006, 15 (3): 229233.
- 相似文献/References:
-
[1]赵春晖,肖晓俊.克隆粒子群算法的镜像层叠滤波器的优化设计[J].智能系统学报,2008,3(2):39.
备注/Memo
收稿日期：2007-09-20.
基金项目：国家自然科学基金资助项目(60672034).
作者简介：
赵春晖 男，1965年生，教授、博士生导师，主要研究方向为智能信息处理技术、图像处理 .获省部级科技奖5项，发表论文180多篇，出版著作3部.
通讯作者：赵春晖.zhaochunhui@hrbeu.edu.cn.
更新日期/Last Update:
2009-05-11