| 摘要: |
| 针对空战目标识别中机型自动识别比较困难的问题，提出了采用航迹特征的智能目标识别方法。利用卷积神经网络（Convolutional Neural Network，CNN）分层学习特征的能力，训练CNN算法模型自动地从航迹数据中学习有用的特征并分类。利用沿海实地采集的15个类别的飞机航迹数据，经一系列数据预处理后作为智能识别算法的训练和测试数据，在验证实验中描述了算法网络的相关配置,对比了CNN与其他分类器的识别结果。实验结果表明，CNN具有很好的识别性能。 |
| 关键词: 目标识别 机型识别 航迹特征 卷积神经网络 |
| DOI： |
|
| 基金项目: |
|
| Intelligent aircraft type recognition based on flight trajectory features |
| XU Xiong |
| ((Southwest China Institute of Electronic Technology,Chengdu 610036,China)) |
| Abstract: |
| To solve the problem of aircraft type automatic recognition in air combat target recognition,an intelligent target recognition method based on trajectory characteristics is proposed.The convolutional neural network (CNN) is used to train the CNN algorithm model to automatically learn useful features from trajectory data and classify them.Fifteen kinds of airplane trajectory data collected from coastal areas are used as training and testing data of intelligent recognition algorithm after a series of data preprocessing.The related configuration of the algorithm network is described in the validation experiment,and the recognition results of CNN and other classifiers are compared.The experimental results show that CNN has good recognition performance. |
| Key words: target recognition aircraft type recognition trajectory feature convolutional neural network |
《电讯技术》编辑部 