A method based on ICA and multiple VQ for acoustic target identification

Author

Zhang, Jianqiang ; Yang, Hongmei ; Li, Yaobo ; Liu, Enyang ; Chen, Gong

Author_Institution

Electron. Eng. Coll., Naval Univ. of Eng., Wuhan

fYear

2008

fDate

20-23 June 2008

Firstpage

1650

Lastpage

1655

Abstract

An identification method based on multiple template vector quantization is proposed through extracting LPC characteristic from acoustic target under noise surroundings. This paper introduces an approach for training acoustic target codebooks in the battlefield using multiple session training samples with every target having multiple codebooks. Finally target is identified by vector quantization. Simulation indicates this method is effective in noisy target identification.

Keywords

acoustic signal processing; independent component analysis; linear predictive coding; target tracking; vector quantisation; LPC characteristic; acoustic target codebook; acoustic target identification; independent component analysis; multiple template vector quantization; noise surroundings; noisy target identification; Acoustic noise; Acoustic waves; Acoustical engineering; Degradation; Higher order statistics; Independent component analysis; Linear predictive coding; Predictive models; Target recognition; Vector quantization; LPC; multiple template vector quantization; simulation; target recognition;

fLanguage

English

Publisher

ieee

Conference_Titel

Information and Automation, 2008. ICIA 2008. International Conference on

Conference_Location

Changsha

Print_ISBN

978-1-4244-2183-1

Electronic_ISBN

978-1-4244-2184-8

Type

conf

DOI

10.1109/ICINFA.2008.4608269

Filename

4608269