基于低分辨率雷达的海面舰船目标分类识别技术  

Ship Target Classification Based on the Low Bandwidth Marine Radar

作　　者：李青[1] 李斌[1] 胡文俊[1] 赵春华[1] 刘蕾蕾[1] 

机构地区：[1]南京电子技术研究所,南京210039

出　　处：《现代雷达》2012年第12期45-49,共5页Modern Radar

摘　　要：文中基于雷达窄带信息提出了一种对海面舰船进行大小及吨位进行目标分类识别的算法。算法通过傅里叶-梅林变换及Fisher判别分析提取分类特征,采用支撑矢量机进行分类识别,最后利用实测的机载雷达数据验证了此方法的有效性。该分类识别算法能够在不影响雷达正常警戒工作任务的前提下,对视场内所有探测到的海面目标进行分类辨识,在工程上具备较高的实用意义。Abstract ：A new method for marine ships＇ recognition and classification based on the low resolution radar is proposed in this paper. The method can classify ships into different size and tonnage according the to narrowband echoes of the target. In the paper, the practical experiment of the airborne radar data has been done to validate the performance of proposed algorithm. The advantage of the method is that it can be embedded into the detection work mode （ the normal mode of the radar） and can classify all the ships properties detected by the radar.

关 键 词：舰船分类 傅里叶-梅林变换 FISHER判别分析 支撑矢量机 模式识别 

分 类 号：TN957.51[电子电信—信号与信息处理]