首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于渔探仪的目标双谱特征提取
引用本文:于铭,杨士莪,陈楚.基于渔探仪的目标双谱特征提取[J].山东农业大学学报(自然科学版),2005,36(3):367-369,372.
作者姓名:于铭  杨士莪  陈楚
作者单位:哈尔滨工程大学水声工程学院,黑龙江,哈尔滨,150001
基金项目:科技部基础部技术工作项目资助,项目编号2001DEA1004
摘    要:本文研究了基于双谱估计的渔探仪水下目标信号的特征提取方法。利用高阶谱来提取水下目标特征能有效的抑制信号的高斯噪声,不受假设目标信号为高斯分布的限制,并且能有效提取渔探仪所采集到的水下目标信号的非高斯特征分量。在提取了水下目标信号的双谱特征的基础上,利用BP神经网络对五类水下目标信号的双谱特征进行分类,得到了较高的识别率。

关 键 词:渔探仪  双谱  特征提取  目标识别
文章编号:1000-2324(2005)03-0367-03
收稿时间:2004-10-20
修稿时间:2004-10-20

FEATURE EXTRACTION BY BISPECTRUM OF FISHERY DETECTOR TARGET
YU Ming,YANG Shi-e,CHEN Chu.FEATURE EXTRACTION BY BISPECTRUM OF FISHERY DETECTOR TARGET[J].Journal of Shandong Agricultural University,2005,36(3):367-369,372.
Authors:YU Ming  YANG Shi-e  CHEN Chu
Institution:College of Underwater Acoustic Engineering, Harbin Engineering University, Harbin 150001, China
Abstract:In this paper, the method of extract the feature of target signal underwater is studied using bispectrumestimation. Using high order spectrum to extract target feature underwater can restrain the Gaussian noise, and at the same time it can attain the no Gaussian feature of target without Gaussian hypothesis. Classifying five target signals underwater base on the bispectrum features through BP neural networks, the targets are identified properly with high efficiently.
Keywords:fishery detector  bispectrum  feature extraction  target identification
本文献已被 CNKI 维普 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