Answers are grounded in this paper and cite their sources.
Aerial target recognition based on inverse synthetic aperture radar imaging is a critical technology in military reconnaissance, yet it is severely challenged by the attitude sensitivity of electromagnetic scattering. Due to the noncooperative nature of targets, it is difficult to obtain a complete feature template library, which degrades recognition reliability. This article proposes a novel recognition method under incomplete sample conditions. First, a scattering center evolution model (SCEM) is developed to describe the attitude-dependent variation of dominant scattering centers in a physically constrained form. Unlike using isolated scattering centers, the SCEM couples endpoint and edge features, introducing attitude visibility and low-order evolution functions to parameterize amplitude variations. Based on sparse observations, SCEM parameters are estimated to reconstruct attitude-controllable feature templates. Furthermore, the reconstructed space is embedded into a discriminative linear subspace, where Fisher-type selection and a similarity classifier are adopted for recognition. Experiments on electromagnetic calculation and measured data confirm that the proposed method outperforms template averaging and implicit neural representation in both reconstruction similarity and recognition accuracy.
No takes yet. Share an insight, caveat, or question.
Wu et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: