离散贝叶斯分类算法雷达目标一维距离像识别  

Range profile-based radar target recognition using discretized Bayes classifying algorithm

作　　者：周德全[1] 郭艳清[1] 

机构地区：[1]中原工学院电子信息学院,郑州451191

出　　处：《电光与控制》2007年第5期79-81,共3页Electronics Optics & Control

摘　　要：首先讨论了传统贝叶斯分类器设计方法的局限性,然后提出了离散贝叶斯分类算法,并推导了离散贝叶斯分类器的分类误差估算公式。用大量的实测高分辨雷达目标回波数据进行计算机模拟,离散贝叶斯分类算法获得了平均94.33%的正确识别率,分类误差估算公式估算的分类误差为5.23%。计算机模拟结果表明了该方法的有效性。The limitation of traditional designing methods of Bayes classifier is discussed at first. Then the algorithm of discrete Bayes classifier is proposed, and formulas for estimating classifying error of Bayes classifier are deduced. A computer simulation was carried out by using a large quantity of measured high-resolution radar return data of the target, an average correct identification rate of 94.33 % and estimating classifying error of 5.23 % were obtained. The results showed that the algorithm and formulas are correct and effective.

关 键 词：贝叶斯 分类器 雷达目标识别 一维距离像 误差 

分 类 号：V271.4[航空宇航科学与技术—飞行器设计] TN957.51[电子电信—信号与信息处理]