Feature extraction and fusion based on the characteristics of underwater targets

被引：0
|
作者：
Li X.-K. [1 ]
Li T.-T. [1 ]
Xia Z. [1 ]
机构：
[1] Acoustic Science and Technology Laboratory, Harbin Engineering University
来源：
Harbin Gongcheng Daxue Xuebao/Journal of Harbin Engineering University | 2010年 / 31卷 / 07期
关键词：
FDWT; FRFT; Hilbert marginal spectrum; Hilbert spectrum; Underwater target recognition;
D O I：
10.3969/j.issn.1006-7043.2010.07.014
中图分类号：
学科分类号：
摘要：
In order to solve the underwater targets recognition problem based on the echo analysis method, the time-frequency characteristic, multi-components characteristic, energy integral characteristic of the targets echo are discussed in this paper. The four kinds of time-frequency analysis methods including frequency discrete wavelet transform (FDWT), Hilbert spectrum, Hilbert marginal spectrum, fractional Fourier transform (FRFT) take different aspects into the characteristics of the targets echo and reverberation. The extracted features are compressed and fused before sending to the support vector machine to make an identification of the target echo and reverberation. The recognition rates of time-frequency features and fusion feature dealing with the experiment data got by transmit-received sonar of different inclination angles are given. The results showed that the recognition rate is higher when the inclination angle is larger. The fusion method can effectively improve the recognition rate.
引用
收藏
页码：903 / 908
页数：5
相关论文
共 12 条
  • [1] Peine H., Brecht D., Broder F., Detection of objects buried in the seafloor, Acta Acustica United with Acustica, 92, pp. 150-152, (2006)
  • [2] Sternlicht D.D., Lemonds D.W., Dikeman R.D., Et al., Detection and classification of buried objects with an adaptive acoustic mine-hunting system, Oceans Conference Record, 1, pp. 212-219, (2001)
  • [3] Yoder T.J., Bucaro J.A., Houston B.H., Simpson H.J., Long range detection and identification of underwater mines using very low frequencies (1-10 kHz), Proceedings of SPIE-The International Society for Optical Engineering, 3392, pp. 203-210, (1998)
  • [4] Robinson M., Azimi-Sadjadi M.R., Sternlicht D.D., Lemonds D., Multi-aspect acoustic classification of buried objects, Proc of OCEANS, pp. 478-484, (2003)
  • [5] Tang W., Highlight model of echoes from sonar targets, Acta Acustica, 19, 2, pp. 92-100, (1994)
  • [6] pp. 110-124, (2001)
  • [7] Li X., Study on characteristic extraction and identification of mine targets, pp. 58-66, (2000)
  • [8] Huang N.E., The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis, J Proc R Soc London A, 454, pp. 903-995, (1998)
  • [9] pp. 23-35, (2004)
  • [10] Schulkopf B., Smola A., Muller K.R., Kernel principal component analysis, Proc of ICANN'97, pp. 583-589, (1997)
12