华中科技大学学报：自然科学版2012，Vol.40Issue(9)：25-29,5.
模糊聚类的侧扫声纳图像分割算法
Segmentation algorithm of fuzzy clustering on side scan sonar image
王雷 1叶秀芬 1王天1
作者信息
- 1. 哈尔滨工程大学自动化学院,黑龙江哈尔滨150001
- 折叠
摘要
Abstract
Aiming at the characteristics of background areas in the sonar imagery,the side scan sonar image was decomposed into several intrinsic mode functions(IMFs) and a residue with the bidimensional empirical mode decomposition(BEMD).Features of the targets and shadows were enhanced and the noise was reduced.In order to use the spatial relationships between pixels,Gauss Markov random field(GMRF) was used to extract the local texture,and the mis-segmentation was reduced.Finally,a new fuzzy clustering algorithm was proposed,and its clustering criterion and distance function were rewritten.By the comparison with other traditional clustering methods in sonar image segmentation field,the proposed algorithm was proved to be superior to others.关键词
声纳图像/频率/模糊聚类/二维经验模态分解（BEMD）/固有模态函数（IMF）/高斯马尔可夫纹理（GMRF）Key words
sonar image/frequency/fuzzy clustering/bidimensional empirical mode decomposition(BEMD)/intrinsic mode functions(IMFs)/Gauss Markov random field(GMRF)分类
信息技术与安全科学引用本文复制引用
王雷,叶秀芬,王天..模糊聚类的侧扫声纳图像分割算法[J].华中科技大学学报：自然科学版,2012,40(9):25-29,5.
