ISAR target micro-Doppler feature extraction and recognition method based on PCA and ICA

摘要

Micro-Doppler features can be regarded as a unique signature of ISAR targets and provide critical information for target recognition. This paper proposes an ISAR micro-Doppler feature extraction and recognition method based on combination algorithm of principle component analysis (PCA) and independent component analysis (ICA). In order to decrease computation burden, PCA algorithm is first applied to reduce dimension and second order correlation, and then ICA decomposes micro-Doppler represented in time frequency domain into a set of independent basis components, which consist the micro-Doppler feature subspace. Target recognition is transformed to calculate the projection coefficient in the subspace. The experimental results verify the correctness and effectiveness of the proposed method.