Abstract
It is important for Inverse Synthetic Aperture Radar (ISAR) to improve its resolution, which can get more critical information. In this paper, an ISAR super-resolution imaging algorithm based on joint dictionary learning is proposed, which means we find two special sets of sparse signals from numerous high and low resolution ISAR images respectively, this two sets of sparse elements which are also called dictionaries can represent all the original signal linearly. Through the learned dictionaries, we can achieve the super resolution of ISAR images. A coupled dictionary learning algorithm based on Restricted Boltzmann Machine (RBM) is designed to learn high and lowresolution dictionaries. Finally, low-resolution ISAR images are sparsely reconstructed based on dictionaries. The suggested approach has lower computational complexity and higher anti-noise performance.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have