基于Transformer的舰船高分辨距离像目标识别  

Target Recognition of Ship High Resolution Range Profile Based on Transformer

作　　者：但波 宋伟健 薛永华 卢中原 DAN Bo;SONG Weijian;XUE Yonghua;LU Zhongyuan(Naval Aviation University,Yantai 264001;Yantai University,Yantai 264001)

机构地区：[1]海军航空大学,烟台264001 [2]烟台大学,烟台264001

出　　处：《舰船电子工程》2024年第7期25-28,共4页Ship Electronic Engineering

摘　　要：基于Transformer模型出色的序列数据相关性保持以及捕获全局特征的能力,提出了一种新的基于Transformer编码器的建模方法用于舰船识别。该方法首先利用编码器部分对舰船高分辨距离像数据进行建模,然后使用位置编码来记录不同序列之间的位置信息,最后利用自注意力机制对整个舰船高分辨距离像序列进行建模,通过实验数据表明,所提出的模型识别准确率达到95.98%,具有出色的识别性能。A new modeling method based on Transformer encoders is proposed for ship recognition based on the Transformer model's excellent serial-data correlation preserving ability and the ability to capture global features.Firstly,the encoder is used to model the ship's high resolution range profile data,and then position coding is used to record the position information between differ⁃ent sequences.Finally,the self-attention mechanism is used to model the whole ship's high resolution range profile sequence.The experimental data show that the recognition accuracy of the proposed model is 95.98%.It has excellent recognition performance.

关 键 词：TRANSFORMER 舰船识别 高分辨距离像 自注意力机制 

分 类 号：O235[理学—运筹学与控制论]