Answers are grounded in this paper and cite their sources.
Modern radar systems are expected not only to indicate the presence of an object but also to provide Automatic Target Recognition and Non-Cooperative Target Recognition. It seems appealing to apply state-of-the-art Artificial Intelligence algorithms for this purpose, but signature variability of 2D ISAR and limited data availability makes classifier training difficult. The usage of 3D InISAR is intended to lessen angle dependency of the formed images, and this way improve classification accuracy and limit time of training for a given number of training samples. In this paper, we present a 3D InISAR classification method based on the simple Convolutional Neural Network (CNN). We compare it to the 2D ISAR CNN classification and show that it enhances both training efficiency and classification accuracy.
No takes yet. Share an insight, caveat, or question.
Wielgo et al. (2022) studied this question.
Synapse has enriched 8 closely related papers on similar clinical questions. Consider them for comparative context: