基于ISAR像的舰船目标识别技术研究  

A Study on Ship Classification Based on ISAR Images

作　　者：侯颖妮[1] 杨予昊[1] 李士国[1] 江涛[1] 

机构地区：[1]南京电子技术研究所,南京210039

出　　处：《现代雷达》2016年第3期47-50,53,共5页Modern Radar

摘　　要：针对海面舰船目标自动识别问题,在舰船目标逆合成孔径雷达(ISAR)图像预处理与结构特征提取的基础上,研究了采用ISAR轮廓像与舰船模型轮廓像匹配进行目标识别的方法。该方法首先利用结构特征进行粗分类,缩小模型匹配范围,接着将模型库中的模型投影到距离多普勒平面,利用二者轮廓像的相似程度完成目标识别。雷达实测数据处理说明了该方法的有效性。For the radar auto classification of ship, a method based on ISAR silhouette and ship model silhouette matching is studied. First the coarse classification is implemented on the base of ISAR image pre-processing and structure features extraction, in order to reduce the number of candidate models, and then the candidate models are projected to range-Doppler domain. The ship is classified according to the best match of the silhouettes. Last the validly of the method is demonstrated by the real radar data processing.

关 键 词：舰船ISAR 目标识别 轮廓匹配 特征提取 

分 类 号：TN957.52[电子电信—信号与信息处理]