CAJ | 학술논문

提出了利用支持向量机对战场侦察雷达目标回波信号进行处理,以实现对卡车、坦克等在地面运动的车辆目标进行分类识别的一种新算法.首先对雷达接收到的目标回波信号作频域分析,从中提取待分类目标信号的特征向量,然后利用所建立的支持向量机模型对目标信号作训练和识别,最后与经典谱分析和神经网络的方法作比较,并采用实际数据验证这种识别方法的有效性.
제출료이용지지향량궤대전장정찰뢰체목표회파신호진행처리,이실현대잡차、탄극등재지면운동적차량목표진행분류식별적일충신산법.수선대뢰체접수도적목표회파신호작빈역분석,종중제취대분류목표신호적특정향량,연후이용소건립적지지향량궤모형대목표신호작훈련화식별,최후여경전보분석화신경망락적방법작비교,병채용실제수거험증저충식별방법적유효성.
A new method of target recognition is put forward, in which echo of moving targets, such as truck and tank, detected by the scout radar is processed by means of support vector machine (SVM). Firstly, by analyzing echo of targets received by radar in frequency domain, eigenvector of the signal is extracted, then SVM model is used to train and recognize these targets. Finally, by comparing with recognition method of classic spectrum analysis and neural network, this new method is proved to be efficient in virtue of real data.