首页 | 本学科首页   官方微博 | 高级检索  
     检索      

星载合成孔径雷达图像的飞机目标检测
引用本文:郭倩,王海鹏,徐丰.星载合成孔径雷达图像的飞机目标检测[J].上海航天,2018(6):57-64.
作者姓名:郭倩  王海鹏  徐丰
作者单位:复旦大学 电磁波信息科学教育部重点实验室,上海 200433,复旦大学 电磁波信息科学教育部重点实验室,上海 200433,复旦大学 电磁波信息科学教育部重点实验室,上海 200433
基金项目:国家自然科学基金(61571132);上海航天科技创新基金(SAST2017078,SAST2016061)
摘    要:针对大场景下星载合成孔径雷达(SAR)图像中飞机目标检测问题,提出一种端到端的飞机目标检测算法。先在大场景SAR图像中对机场目标进行粗检测,定位机场区域,再通过精确分割算法获得机场的精细区域。对机场区域中的飞机目标进行检测,采用一种基于Canny算子的边缘检测与卷积神经网络结合的飞机目标检测算法。通过飞机边缘检测、边界框预处理等操作确定潜在飞机目标在机场中的位置范围,采用基于GoogLeNet的卷积神经网络对可疑目标进行鉴别。利用星载合成孔径雷达数据对算法进行验证,证明该方法的有效性与实用性。

关 键 词:合成孔径雷达    替代滤波    飞机检测    边缘检测    卷积神经网络
收稿时间:2018/10/8 0:00:00
修稿时间:2018/10/25 0:00:00

Aircraft Target Detection from Spaceborne Synthetic Aperture Radar Image
GUO Qian,WANG Haipeng and XU Feng.Aircraft Target Detection from Spaceborne Synthetic Aperture Radar Image[J].Aerospace Shanghai,2018(6):57-64.
Authors:GUO Qian  WANG Haipeng and XU Feng
Institution:Key Laboratory for Information Science of Electromagnetic Waves, Fudan University, Shanghai 200433, China,Key Laboratory for Information Science of Electromagnetic Waves, Fudan University, Shanghai 200433, China and Key Laboratory for Information Science of Electromagnetic Waves, Fudan University, Shanghai 200433, China
Abstract:Aiming at the aircraft target detection in big scenes from spaceborne synthetic aperture radar (SAR) image, an end-to-end algorithm is proposed in this paper. It firstly locates the airport area from the large SAR image via rough detection method, and then obtains an accurate mask of the airport by using accurate segmentation method. Within the airport area, aircraft targets are detected via a two-stage method. In the first stage, potential targets are identified based on Canny edge-detection, and in the second stage, true aircraft targets are discriminated via GoogLeNet convolutional neural network. The effectiveness and practicality of the method are verified by collected spaceborne SAR image data.
Keywords:synthetic aperture radar(SAR)  alternative filter  aircraft detection  edge detection  convolutional neural network
本文献已被 CNKI 等数据库收录!
点击此处可从《上海航天》浏览原始摘要信息
点击此处可从《上海航天》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号