基于深度学习的复杂气象条件下海上船只检测

被引:11
作者
熊咏平 [1 ,2 ]
丁胜 [1 ,2 ]
邓春华 [1 ,2 ]
方国康 [1 ,2 ]
龚锐 [1 ,2 ]
机构
[1] 武汉科技大学计算机科学与技术学院
[2] 智能信息处理与实时工业系统湖北省重点实验室
关键词
YOLO v2; 目标检测; 多尺度目标检测; 显著性分割;
D O I
暂无
中图分类号
TP18 [人工智能理论]; TP751 [图像处理方法];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ; 081002 ;
摘要
为了解决复杂海情环境下的不同种类和大小的舰船检测问题,提出一种实时的深度学习的目标检测算法。首先,提出了一种清晰图片和模糊图片(雨、雾等图片)判别的方法;然后,在YOLO v2的深度学习框架的基础上提出一种多尺度目标检测算法;最后,针对遥感图像舰船目标的特点,提出了一种改进的非极大值抑制和显著性分割算法,对最终的检测结果进一步优化。在复杂海情和气象条件下的舰船目标公开比赛的数据集上,实验结果表明,相比原始的YOLO v2,该方法的准确率提升了16%。
引用
收藏
页码:3631 / 3637
页数:7
相关论文
共 20 条
[1]  
Deep learning. Lecun Y,Bengio Y,Hinton G. Nature . 2015
[2]  
Cnn Features off-the-shelf:An Astounding Baseline for Recognition. Razavian A S,Azizpour H,Sullivan J,et al. Computer Vision and Pattern Recognition Workshops . 2014
[3]  
Rich feature hierarchies for accurate object detection and semantic segmentation. Girshick R,Donahue J,Darrell T,et al. Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition CVPR . 2014
[4]  
Spatial pyramid pooling in deep convolutional networks for visual recognition. He K,Zhang X,Ren S, et al. Pattern Analysis and Machine Intelligence, IEEE Transactions on . 2015
[5]  
Faster R-CNN:towards real-time object detection with region proposal networks. Ren S,He K,Girshick R,et al. International Conference on Neural Information Processing Systems . 2015
[6]  
You only look once:Unified,real-time object detection. REDMON J,DIVVALA S,GIRSHICK R,et al. IEEE Conference on Computer Vision and Pattern Recognition (CVPR) . 2015
[7]  
Chinese Marine Economy Development: Dynamic Evolution and Spatial Difference[J]. SUN Caizhi,LI Xin,ZOU Wei,WANG Song,WANG Zeyu.  Chinese Geographical Science. 2018(01)
[8]   Distinctive image features from scale-invariant keypoints [J].
Lowe, DG .
INTERNATIONAL JOURNAL OF COMPUTER VISION, 2004, 60 (02) :91-110
[9]  
SSD:single shot multibox detector. LIU W,ANGUELOV D,ERHAN D,et al. Proceedings of the 14th European Conference on Computer Vision . 2015
[10]  
Fast R-CNN. GIRSHICK R. http://cn.arxiv.org/pdf/1504.08083.pdf . 2018