首页 | 本学科首页   官方微博 | 高级检索  
     

基于BEMD的水下声影区地貌及纹理特征提取
引用本文:葛光涛,桑恩方,刘卓夫,吴玉桐. 基于BEMD的水下声影区地貌及纹理特征提取[J]. 郑州大学学报(工学版), 2007, 28(4): 70-74
作者姓名:葛光涛  桑恩方  刘卓夫  吴玉桐
作者单位:哈尔滨工程大学,水声工程学院,黑龙江,哈尔滨150001
摘    要:提出了一种基于二维经验模态分解(The Bidimensional Empirical Mode Decomposition,BEMD)的水声图像声影区地貌及纹理特征提取方法.为了解决声影区的存在对水声图像特征提取的干扰问题,先对水声图像进行二维模态分解,采用Canny边缘检测器对其中的模态1或几个模态的叠加进行特征提取.实验结果表明,该方法增强了声影区内目标的特征信息,弱化了声影区的阴影边缘,是一种实用的水声图像特征提取新方法.

关 键 词:声影区  特征提取  边缘检测  二维经验模态分解
文章编号:1671-6833(2007)04-0070-05
修稿时间:2007-08-18

Feature Extraction of Physiognomy and Texture in the Underwater Acoustic Shadow Field Based on Bidimensional Empirical Mode Decomposition
GE Guang-tao,SANG En-fang,LIU Zhuo-fu,WU Yu-tong. Feature Extraction of Physiognomy and Texture in the Underwater Acoustic Shadow Field Based on Bidimensional Empirical Mode Decomposition[J]. Journal of Zhengzhou University: Eng Sci, 2007, 28(4): 70-74
Authors:GE Guang-tao  SANG En-fang  LIU Zhuo-fu  WU Yu-tong
Abstract:A method for the underwater acoustic shadow field feature extraction of physiognomy and texture based on the Bidimensional Empirical Mode Decomposition(BEMD) has been introduced in this paper.To reduce the disturbance from the acoustic shadow field,the author decomposes the underwater acoustic image into several IMFs and a residue,then better features from the first IMF or the sum of several IMFs with the Canny edge detector can be extracted.Experiments prove that this new method really enhances the features of objects in the acoustic shadow field and weakens the edge of the acoustic shadow field.
Keywords:acoustic shadow field  character extraction  edge detection  bidimensional empirical mode decomposition(BEMD)
本文献已被 CNKI 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号