1. Science
  2. Publications
  3. Information Processing Systems
  4. 2(92)'2011
  5. Radar objects classification by using information features containing in backscattering signal time-frequency distributions

Radar objects classification by using information features containing in backscattering signal time-frequency distributions

A.V. Totsky
Annotations languages:

Novel approach to extraction of information features based on backscattering signal parameters estimation is proposed for solving moving object classification problem. It has been proposed to extract information features from short-time echo-signal bispectral estimates measured by Doppler ground radar surveillance system. Results of experimental study of automatic radar systems are represented for surveillance the moving ground objects. Comparative analysis is performed for radar pattern classification probabilities by using proposed bicepstral coefficients and common cepstral coeffisients.
Keywords: radiolocation, radar objects classification, information features, bispectrum, cepstrum, bicepstral coefficients