基于空中目标识别的特征提取与选择  

Features Extraction and Selection of Air Target Recognitions

作　　者：田瑞娟[1] 杨帆 

机构地区：[1]中国兵器工业第五八研究所军品部,四川绵阳621000 [2]总装备部重庆军代局,重庆400060

出　　处：《兵工自动化》2014年第3期80-83,共4页Ordnance Industry Automation

摘　　要：为从冗杂的海量信息中提取出有效的特征数据和信息,对空中目标识别的目标特征进行研究。介绍目标跟踪识别系统,从搜索雷达的目标回波信号特征、目标光谱成像特征、目标姿态与运动特征、目标特征选择与融合等方面进行目标识别特征提取,生成图像与特征数据库,并对选取的目标特征进行多特征融合的识别精度验证。结果表明:提取与选择的目标特征在识别中取得较满意效果,能证明选取的空中目标特征在目标识别算法中的可靠性。For extracting valid feature data and information from massive data, research on target feature of air target recognition. Introduce target tracing recognition system, carry out target recognition feature extraction from searching radar target echo signal feature, target spectral imaging feature, target gesture, motion feature, target feature selection and fusion. Create image and feature data base, and carry out multi-features fusion recognition accuracy validation for selected target feature. The results show that the extracted and selected target feature received great effect, the reliability of selected air target feature in target recognition method is proved.

关 键 词：空中目标 特征提取 识别 

分 类 号：TJ03[兵器科学与技术—兵器发射理论与技术]