A Review of SAR Images Target Recognition Based on Few-shot

Abstract:
Synthetic Aperture Radar (SAR) image target recognition based on deep learning has become one of the key technologies in the field of radar detection, and the lack of SAR data sets has limited the research development of deep neural networks that rely on a large number of training samples. In this paper, we firstly introduce the SAR image data expansion methods, secondly provide an in-depth review of the current status of the migration learning-based small sample SAR target recognition methods for the small sample recognition problem, and finally, focus on some key problems that need to be solved in the field of small sample-based SAR image target recognition, and provide a long-term outlook on the development of these problems.