Multidimensional Radar Imaging for Target Identification

Author(s): Alessandro Vetere; Eugenio Pino; Adriano Meta  

DOI: 10.14339/STO-SET-273-21     |     ISSN: TBD  

Automatic Target Recognition (ATR) with Artificial Intelligence (AI) plays a key role in modern surveillance and reconnaissance activities. The enabling technology for ATR is the availability of a database that is representative of realistic situations including targets and backgrounds. When applied to radar imaging the problem of collecting enough data of targets under different configurations and conditions becomes extremely complicated and unaffordable

Citation:

Vetere A.; Pino E.; Meta A.: Physics-Based GPU-Accelerated SAR Simulator for ATR Database Generation. Multidimensional Radar Imaging for Target Identification, NATO STO Review, Spring 2022.

ABSTRACT

Automatic Target Recognition (ATR) with Artificial Intelligence (AI) plays a key role in modern surveillance and reconnaissance activities. The enabling technology for ATR is the availability of a database that is representative of realistic situations including targets and backgrounds. When applied to radar imaging the problem of collecting enough data of targets under different configurations and conditions becomes extremely complicated and unaffordable. At MetaSensing we are tackling this problem with a recently developed physics-based SAR simulator (KAISAR) able to estimate the complex 3D RCS (Radar Cross Section) of a target and to model the SAR collection and processing stages to build a synthetic yet realistic database of target-background-images with data augmentation. This paper provides a description of the SAR simulator and some examples of synthetic data generated.

DOWNLOAD FULL ARTICLE


Activity Type: Specialists' Meeting
Reference: STO-SET-273
Date:

SET-273 Papers