An Effective Deconvolution Algorithm for Time-Domain Near-Field ISAR Imaging

In this paper, we present a deconvolution algorithm for time-domain near-field ISAR imaging. Due to its impulse-like waveform, the backscattering waveform from an electrically large conducting sphere is adopted as the system response to restore the impulse response, and to reconstruct the images. The deconvolution is discretized into a matrix equation, and the least-squares solution is implemented by the conjugate gradient method. Furthermore, the Coifman wavelet (Coiflets) of the 6th order is employed to replace the system function of the 1st and 3rd derivatives of a Gaussian function. The resulting images are clean with most smearing removed. Owing to the orthogonality and compact support of the Coiflets, the deconvolution process is accelerated 22 times without sacrificing the image quality