Medical & Biological Engineering & Computing

, Volume 45, Issue 6, pp 611–616

Enhanced T-ray signal classification using wavelet preprocessing

  • X. X. Yin
  • K. M. Kong
  • J. W. Lim
  • B. W.-H. Ng
  • B. Ferguson
  • S. P. Mickan
  • D. Abbott
Short Communication

DOI: 10.1007/s11517-007-0185-y

Cite this article as:
Yin, X.X., Kong, K.M., Lim, J.W. et al. Med Bio Eng Comput (2007) 45: 611. doi:10.1007/s11517-007-0185-y

Abstract

This study demonstrates the application of one-dimensional discrete wavelet transforms in the classification of T-ray pulsed signals. Fast Fourier transforms (FFTs) are used as a feature extraction tool and a Mahalanobis distance classifier is employed for classification. Soft threshold wavelet shrinkage de-noising is used and plays an important role in de-noising and reconstruction of T-ray pulsed signals. An iterative algorithm is applied to obtain three optimal frequency components and to achieve preferred classification performance.

Keywords

Mahalanobis distance classifier Wavelet denoising T-rays 

Copyright information

© International Federation for Medical and Biological Engineering 2007

Authors and Affiliations

  • X. X. Yin
    • 1
  • K. M. Kong
    • 1
  • J. W. Lim
    • 1
  • B. W.-H. Ng
    • 1
  • B. Ferguson
    • 1
    • 2
  • S. P. Mickan
    • 1
  • D. Abbott
    • 1
  1. 1.Centre for Biomedical Engineering and School of Electrical and Electronic EngineeringThe University of AdelaideAdelaideAustralia
  2. 2.Tenix Defence Systems Pty LtdAdelaideAustralia

Personalised recommendations