摘要
提出了一种基于深度卷积神经网络的遥感图像多类目标识别方法,采用FasterR-CNN模型,以ZFNet为基础网络,通过实验显示对飞机、船舶以及油罐的识别准确率分别为90.67%、93.85%、83.33%,平均识别准确率达到了89.28%.同时与CNN、T-T-RICNN等识别检测方法的识别结果对比发现,提出的FasterR-CNN模型识别性能较好,有着良好的识别检测精度.
In this study,a multi-class target recognition method for remote sensing images based on deep convolutional neural network was proposed. The identification accuracy of aircraft,ships and oil tanks was 90.67%,93.85% and 83.33%,respectively,with an average identification accuracy of 89.28%,using the Faster R-CNN model and ZFNet as the basis network. Meanwhile,compared with the recognition results of CNN,T-T-RICNN and other recognition and detection methods,it is found that the identification performance of the proposed Faster r-cnn model is better,with good recognition and detection accuracy.
作者
胡琼
HU Qiong(School of information and electronic engineering,Liu′an vocational and technical college,Liuan Anhui 237000)
出处
《宁夏师范学院学报》
2019年第10期75-79,共5页
Journal of Ningxia Normal University
基金
高校优秀青年人才计划项目主持(gxyq2019206)
关键词
深度卷积神经网络
遥感图像
多类目标识别
Deep convolutional neural network
Remote sensing images
Multiple target recognition