成喆,吕京国,白颖奇,曹逸飞

• 1:北京建筑大学

摘要(Abstract)：

利用传统方法对遥感影像的目标检测,过程复杂并且耗时。随着深度学习的发展,用深度学习的方法进行目标检测,为遥感影像的检测开辟了新的思路。当前目标检测的方法主要包括以Faster R-CNN为代表的两阶段检测算法和以SSD为代表的单阶段算法,两阶段算法精度高速度慢,单阶段算法速度快精度低。针对两种算法的优势,该文将Faster R-CNN中的RPN与SSD算法相结合,融合单阶段和两阶段算法的优势,在提高精度的情况下保证速度,并加入特征金字塔结构,利用多个卷积层融合低层特征和高层特征的信息,提高预测效果。在NWPUVHR-10高分辨率数据集上进行训练和测试,对结果进行算法评估。同时选用测试集将该文算法与Faster R-CNN和SSD算法进行对比,实验表明该文算法提高了对小目标物体的检测精度,获得了更优的性能。

关键词(KeyWords)： 目标检测;深度学习;RPN网络;SSD算法;遥感影像

Abstract:

Keywords:

基金项目(Foundation): 国家自然科学基金面上项目(41871367,61773377)

作者(Author): 成喆,吕京国,白颖奇,曹逸飞

DOI: 10.16251/j.cnki.1009-2307.2021.04.012

参考文献(References)：

• [1]王彦情，马雷，田原．光学遥感图像舰船目标检测与识别综述[J]．自动化学报，2011,37(9):1029-1039.(WANG Yanqing,MA Lei,TIAN Yuan.State-of-theart of ship detection and recognition in optical remotely sensed imagery[J].Acta Automatica Sinica,2011,37(9):1029-1039．）
• [2]刘扬，付征叶，郑逢斌．高分辨率遥感影像目标分类与识别研究进展[J]．地球信息科学学报，2015,17(9):1080-1091.(LIU Yang,FU Zhengye,ZHENG Fengbin.Review on high resolution remote sensing image classification and recognition[J].Journal of Geoinformation Science,2015,17(9):1080-1091．）
• [3]GIRSHICK R,DONAHUE J,DARRELL T,et al.Rich feature hierarchies for accurate object detection and semantic segmentation[C]∥Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.[S.l.]:IEEE,2014:580-587．
• [4]GU C,LIM J J,ARBELAEZ P,et al.Recognition using regions[C]∥IEEE Conference on Computer Vision&Pattern Recognition.[S.l.]:IEEE,2009:1030-1037．
• [5]HE K,ZHANG X,REN S,et al.Spatial pyramid pooling in deep convolutional networks for visual recognition[J].IEEE Transactions on Pattern Analysis&Machine Intelligence,2014,37 (9):1904-1916．
• [6]GIRSHICK R.Fast R-CNN[C]∥IEEE International Conference on Computer Vision.Washington:IEEEComputer Society,2015:1440-1448．
• [7]REN S,HE K,GIRSHICK R,et al.Faster R-CNN:towards real-time object detection with region proposal networks[C]∥International Conference on Neural Information Processing Systems.Cambridge,Massachusetts:MIT Press,2015:91-99．
• [8]王金传，谭喜成，王召海，等．基于Faster R-CNN深度网络的遥感影像目标识别方法研究[J]．地球信息科学学报，2018,20(10):1500-1508.(WANG Jinchuan,TAN Xicheng,WANG Zhaohai,et al.Faster R-CNNdeep learning network based object recognition of remote sensing image[J].Journal of Geo-information Science,2018,20(10):1500-1508．）
• [9]谢奇芳，姚国清，张猛．基于Faster R-CNN的高分辨率图像目标检测技术[J]．国土资源遥感，2019,31(2):38-43.(XIE Qifang,YAO Guoqing,ZHANG Meng.Research on high resolution image object detection technology based on Faster R-CNN[J].Remote Sensing for Land&Resources,2019,31(2):38-43．）
• [10]HE K,ZHANG X,REN S,et al.Deep residual learning for image recognition[C]∥Proceedings of the IEEEConference on Computer Vision and Pattern Recognition.[S.l.]:IEEE,2016:770-778．
• [11]徐逸之，姚晓婧，李祥，等．基于全卷积网络的高分辨遥感影像目标检测[J]．测绘通报，2018(1):77-82.(XUYizhi,YAO Xiaojing,LI Xiang,et al.Object detection in high resolution remote sensing images based on fully convolution networks[J].Bulletin of Surveying and Mapping,2018(1):77-82．）
• [12]REDMON J,DIVVALA S K,GIRSHICK R,et al.You only look once:unified,real-time object detection[Z/OL].[2020-05-22].https：∥arxiv.org/abs/1506.02640．
• [13]LIU W,ANGUELOV D,ERHAN D,et al.SSD:single shot multibox detector[C]∥European Conference on Computer Vision.Cham:Springer,2016:21-37．
• [14]王俊强，李建胜，周学文，等．改进的SSD算法及其对遥感影像小目标检测性能的分析[J]．光学学报，2019,39(6):373-382.(WANG Junqiang,LI Jiansheng,ZHOUXuewen,et al.Improved SSD algorithm and its performance analysis of small target detection in remote sensing images[J].Acta Optica Sinica,2019,39(6):373-382．）
• [15]SIMONYAN K,ZISSERMAN A.Very deep convolutional networks for large-scale image recognition[C/OL]∥ICLR 2015.[2020-05-22].https：∥arxiv.org/pdf/1409.1556.pdf．
• [16]LIU W,RABINOVICH A,BERG A C.ParseNet:looking wider to see better[C/OL]∥ICLR 2016.[2020-05-22].https：∥arxiv.org/pdf/1506.04579.pdf.