《厦门大学学报（自然科学版）》

针对传统的水声信号分类技术处理方法复杂、特征提取时间长以及特征量多等问题,提出了一种基于稀疏表示的分类系统,先利用正交匹配追踪法(OMP)算法提取与水声信号最为匹配的少数原子作为目标特征,再采用支持向量机(SVM)进行分类.对鲸类声信号进行仿真实验,实验结果表明,不仅提高了压缩效率和运算速度,而且识别率高,在水声信号的实时处理中具有较高的实用价值.

Traditional classification technologies of underwater acoustic signals are involved withissues such as the complicated processing method,the prolonged feature extraction,vast features and other problems.In this paper,we propose a novel method based on sparse representation classification.First,we extract a spot of atoms matched best with underwater acoustic signals as signal features utilizing the OMP algorithm.Second,we adopt SVM as our classifier.Through experimental evaluations,the effect of this method is shown to provide a significant improvement in compression efficiencies,computing speeds and recognition rates.

引言
1 稀疏表示理论
2 基于SVM的分类技术
3 基于稀疏表示的鲸类声信号分类识别
4 结论

图1 蓝鲸声音信号时域波形<br/>Fig.1 Time-domain waveform of the blue-whale sound signal

图1 蓝鲸声音信号时域波形
Fig.1 Time-domain waveform of the blue-whale sound signal

图2 鲸声音样本的稀疏表示系数<br/>Fig.2 Coefficient of sparse representation of whale sounds samples

图2 鲸声音样本的稀疏表示系数
Fig.2 Coefficient of sparse representation of whale sounds samples

表1 分类识别率比较(d=20)<br/>Tab.1 Comparison of classification rate(d=20)

表1 分类识别率比较(d=20)
Tab.1 Comparison of classification rate(d=20)

表2 分类识别性能对比分析(d=4)<br/>Tab.2 Comparison of classification rate(d=4)

表2 分类识别性能对比分析(d=4)
Tab.2 Comparison of classification rate(d=4)

[1] 刘深,张小蓟,牛奕龙,等.基于IMF能量谱的水声信号特征提取与分类[J].计算机工程与应用,2014,50(3):203-206.
[2] 行鸿彦,祁峥东,徐伟.基于选择性支持向量机集成的海杂波背景中的微弱信号检测[J].物理学报,2012,61(24):240504.
[3] DONOHO D L.Compressed sensing [J].IEEE Transactions on Information Theory,2006,52(4):1289-1306.
[4] CANDÈS E J,ROMBERG J,TAO T.Robust uncertainty principles:exact signal reconstruction from highly incomplete frequency information[J].IEEE Transactions on Information Theory,2006,52(2):489-509.
[5] VAPNIK V.The nature of statistical leaming theory[M].New York:Springer-Verlag,1995:76-79.
[6] 李欣新.船舶及鲸类声信号特征提取和分类识别研究[D].哈尔滨:哈尔滨工程大学,2012:72-74,77-78.
[7] 乔钢,王巍,王玥,等.基于压缩感知的OFDM水声通信信道二次估计算法[J].声学技术,2013,32(5):357-361.
[8] CHEN S S,DONOHO D L,SAUNDERS M A.Atomic decomposition by basis pursuit[J].SIAM Journal on Scientific Computing,1999,20(1):33-61.
[9] NEEDELL D,VERSHYNIN R.Greedy signal recovery and uncertainty principles[C]∥Computational Imaging Ⅵ.San Jose:SPIE,2008:68140J.doi :10.1117/12.776996.
[10] TROPP J A,GILBERT A C.Signal recovery form random measurements via orthogonal matching pursuit [J].IEEE Transactions on Information Theory,2007,53(12):4655-4666.
[11] VAPNIK V N.Statistical learning theory [M].New York:John Wiley and Sons,1998:81-85.
[12] CORTES C,VAPNIK V N.Support vector networks [J].Machine Learning,1995,20(3):273-297.
[13] 江伟华,曹秀岭,童峰.采用支持向量机的水声通信信号调制识别方法[J].厦门大学学报(自然科学版),2015,54(4):534-539.
[14] 韩笑,殷敬伟,郭龙祥,等.基于差分 Pattern 时延差编码和海豚 whistles 信号的仿生水声通信技术研究[J].物理学报,2013,62(22):244-249.
[15] MARK A M.Whale acoustics[EB/OL].[2016-03-01].http:∥www.whaleacoustics.com/audiobaleenwhales.html.

备注

引言

1 稀疏表示理论

2 基于SVM的分类技术

3 基于稀疏表示的鲸类声信号分类识别

4 结论

学报简介

备注

引言

1 稀疏表示理论

2 基于SVM的分类技术

3 基于稀疏表示的鲸类声信号分类识别

4 结 论

学报简介

4 结论