Efficient ISAR Autofocus Technique using Eigenimages

Abstract

In this paper, we propose a new and efficient inverse synthetic aperture radar (ISAR) autofocus technique by introducing eigenimages to boost the speed of the traditional autofocus algorithms. First, a preprocessing step is applied to mitigate the noise components in the received data. Then, we perform an eigen-decomposition of the covariance matrix of the range-aligned data, and generate the signal eigenimage obtained by deriving the Fourier transform of a small number of eigenvectors corresponding to the dominant eigenvalues. Finally, traditional autofocus methods are combined with the proposed signal eigenimage rather than the original ISAR image to eliminate image blurring due to phase errors. The proposed method can significantly lower the computational complexity of the traditional autofocus methods because the dimensionality of the signal eigenimage is considerably smaller than that of the ISAR image. Despite the low dimensionality of the signal eigenimages, the proposed scheme provides well-focused ISAR images that are comparable to those of the traditional autofocus methods in terms of image focal quality. Several simulations and experimental results using measured data of an actual flying aircraft are presented to verify the effectiveness of the proposed method.