基于柔性形态学滤波优化的周期性噪声消除算法  

Periodic noise elimination algorithm based on soft morphological filtering optimization

作　　者：卫星[1] 焦蓬蓬[1] 史永[1] WEI Xing JIAO Pengpeng SHI Yong(Nanjing Normal University Taizhou College, Taizhou 225300, Chin)

机构地区：[1]南京师范大学泰州学院,江苏泰州225300

出　　处：《现代电子技术》2016年第21期70-74,共5页Modern Electronics Technique

基　　金：江苏省高校自然科学基金(15KJB170006);南京师范大学泰州学院院级项目(Q201238)

摘　　要：针对周期性噪声滤波易产生图像失真与降噪效果不佳等问题,提出一种柔性形态学滤波的周期性噪声消除算法。在数学形态学的思想上,构建了一种柔性形态学滤波器,利用形态学开-闭运算和闭-开运算相结合,提高滤波器噪声抑制性能。并利用粒子群优化耦合被动聚集技术,改进信息共享机制,对柔性形态学滤波器的五个主要参数进行优化,输出最优值,从而消除周期性噪声。实验结果表明,与当前降噪技术相比,所提算法对周期性和混合性噪声具有更强的鲁棒性,在消除噪声的同时也较好地保护了图像细节信息。In order to solve the image distortion and poor noise reduction effect of the periodic noise fihering, a periodic noise elimination algorithm based on soft morphological filtering is proposed. On the basis of the mathematical morphology thought, a soft morphological filter was constructed. The combination of morphology open-close operation and close-open opera- tion is used to improve the noise suppression performance of the filter. The technology of using particle swarm optimization to couple the passive congregation is used to improve the information sharing mechanism, optimize the five main parameters of soft morphological filter, output the optimal value, and eliminate the periodic noise. The experimental results show that, in compari- son with the available noise elimination technology, the proposed algorithm has stronger robustness for periodic and mixed noise, and protects the image detail information while removing the noise.

关 键 词：周期性噪声 柔性形态滤波器 粒子群优化 信息共享 被动聚集 

分 类 号：TN911.734[电子电信—通信与信息系统] TP391[电子电信—信息与通信工程]