Moving ground targets classification based on eigenvalue sequence of echo signal

This paper studieds the moving ground targets classification under the circumstances of low resolution and low dwell time. It establishes the echo signal model of micro-motion part using multi-scattering centre model based on micro-Doppler effect, which demonstrated that the number of harmonics contained in the echo signal is a stable feature to distinguish tracked vehicle and wheeled vehicle. Thus a new feature extraction method based on the eigenvalue sequence of the echo signal is proposed, which depicts the harmonic information of the echo signal, and is robust to the change of bulk velocity. The experimental results based on simulated data show the effectiveness of the proposed method and those based on measured data via Support Vector Machine (SVM) show the proposed method outperforms the principle component analysis (PCA) based method in classification performance.