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

舰船辐射噪声的高阶统计量特征提取及特征压缩
引用本文:陈凤林,林正青,彭圆,牟林,张凤珍,王磊.舰船辐射噪声的高阶统计量特征提取及特征压缩[J].应用声学,2010,29(6):466-470.
作者姓名:陈凤林  林正青  彭圆  牟林  张凤珍  王磊
作者单位:水下测控技术国防科技重点实验室
基金项目:水下测控技术国防科技重点实验室基金
摘    要:提取了6类水中目标的11/2谱统计线谱特征和AR模型特征。实验结果表明基于11/2谱的统计线谱特征和AR模型特征有效的识别出了目标,达到了85%的综合识别率。最后利用主元分析技术对上述特征进行了压缩,在保持识别率的同时将特征维数从64维压缩至34维,有利于工程上的实现。

关 键 词:线谱特征  AR模型  主元分析  特征压缩

Extraction and compression of high-order statistical characteristics for ship radiated noise
CHEN Feng-lin,LIN Zheng-qing,PEN Yuan,MOU Lin,ZHANG Feng-zhen and WANG Lei.Extraction and compression of high-order statistical characteristics for ship radiated noise[J].Applied Acoustics,2010,29(6):466-470.
Authors:CHEN Feng-lin  LIN Zheng-qing  PEN Yuan  MOU Lin  ZHANG Feng-zhen and WANG Lei
Institution:(Dalian Scientific Test & Control Technology Institute,Dalian 116013)
Abstract:1 1/2 spectrum statistical line features and the AR model features of six kinds of underwater targets are extracted in this paper.Recognition is done,and the result of which shows that both classes of features are effective in recognizing these targets,and the discrimination is up to 85%.In the last part of this report,a feature compression technology is used to compress the dimension of the original features from 64 to 34,while the discrimination of the compressed features remains unchanged in the main,This could be propitious to engineering realization.
Keywords:Line features  AR model  Principal component analysis  Feature compaction
本文献已被 CNKI 万方数据 等数据库收录!
点击此处可从《应用声学》浏览原始摘要信息
点击此处可从《应用声学》下载免费的PDF全文
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