Space Target Recognition Based on Gramian Angular Field Representation and Attention Mechanism

This paper proposes a method that transforms radar cross section (RCS) data into 2D image through Gramian Angular Field (GAF) transformation and utilizes attention mechanism to achieve recognition of four classical types of space targets. The effectiveness of the proposed method is validated on simulated data, demonstrating noteworthy advancements compared to conventional methods and widely employed deep learning methods.