基于改进型YOLOv3的SAR图像舰船目标检测

doi:10.12305/j.issn.1001-506X.2021.04.10

摘要/Abstract

摘要：

传统合成孔径雷达(synthetic aperture radar, SAR)图像目标检测的方法依赖于人工设计特征且易受复杂背景干扰, 泛化能力较差。深度学习的方法可以自动提取特征且具有良好的抗干扰特性, 对于未来雷达智能感知具有重要意义。不同于其他只能对固定区域进行检测的常规卷积神经网络, 本文提出一种改进型YOLOv3的SAR图像舰船目标检测方法, 该方法基于舰船尺寸与形状自适应采样的可变形卷积、ResNet50变体特征提取器和ShuffleNetv2轻量化思想等设计YOLOv3模型。通过SSDD数据集验证, 在检测效果方面, 相较于原YOLOv3模型, 平均精度从93.21%提高至96.94%, 检测概率从95.51%提高至97.75%;在模型大小方面, 轻量化设计模型仅为原YOLOv3模型的八分之一, 可实现嵌入式的使用。

关键词: 合成孔径雷达, 舰船检测, 单阶段检测算法, 可变形卷积, 卷积神经网络

Abstract:

Traditional synthetic aperture radar(SAR) image target detection methods rely on manual design features and are vulnerable to complex background interference, and their generalization ability is poor. The deep learning method can automatically extract features and has good anti-jamming characteristics, which is of great significance for future radar intelligent perception. Different from other conventional neural networks that can only detect fixed areas, an improved YOLOv3 SAR image ship detection method is proposed in this paper. The new YOLOv3 model is designed based on deforming convolution of adaptive sampling of the ship size and the shape, the ResNet50 variant feature extractor and the ShuffleNetv2 lightweight idea. Through SSDD dataset verification, compared with the original YOLOv3 model, the average accuracy increases from 93.21% to 96.94%, and the detection probability increases from 95.51% to 97.75%. In terms of the model size, the lightweight design model is only one-eighth of the original YOLOv3 model, which can be embedded for use.

Key words: synthetic aperture radar (SAR), ship detection, YOLOv3, deformable convolution, convolutional neural network

中图分类号:

TP751

陈冬, 句彦伟. 基于改进型YOLOv3的SAR图像舰船目标检测[J]. 系统工程与电子技术, 2021, 43(4): 937-943.

Dong CHEN, Yanwei JU. Ship detection in SAR image based on improved YOLOv3[J]. Systems Engineering and Electronics, 2021, 43(4): 937-943.

图/表 10

图1

图2

图3

图4

图5

图6

表1

表2

图7

图8

参考文献 29

1	LI W, ZOU B, XIN Y, et al. An improved CFAR scheme for man-made target detection in high resolution SAR images[C]//Proc. of the IEEE International Geoscience and Remote Sensing Symposium, 2018: 2829-2832.
2	ZHU J , QIU X , PAN Z , et al. Projection shape template-based ship target recognition in TerraSAR-X images[J]. IEEE Geoscience and Remote Sensing Letters, 2017, 14 (2): 222- 226. doi: 10.1109/LGRS.2016.2635699
3	杨国铮, 禹晶, 肖创柏, 等. 基于形态字典学习的复杂背景SAR图像舰船尾迹检测[J]. 自动化学报, 2017, 43 (10): 1713- 1725.
	YANG G Z , YU J , XIAO C B , et al. Ship wake detection in sar images with complex background using morphological dictionary learning[J]. Acta Automatica Sinica, 2017, 43 (10): 1713- 1725.
4	GIRSHICK R, DONAHUE J, DARRELL T, et al. Rich feature hierarchies for accurate object detection and semantic segmentation[C]//Proc. of the Computer Vision and Pattern Recognition, 2014: 580-587.
5	GIRSHICK R. Fast R-CNN[C]//Proc. of the International Conference on Computer Vision, 2015: 1440-1448.
6	REN S , HE K , GIRSHICK R , et al. Faster R-CNN: towards real-time object detection with region proposal networks[J]. IEEE Trans.on Pattern Analysis & Machine Intelligence, 2017, 39 (6): 1137- 1149.
7	LIU W, ANGUELOV D, ERHAN D, et al. SSD: single shot MultiBox detector[C]//Proc. of the European Conference on Computer Vision, 2016: 21-37.
8	LI Z , ZHOU F . FSSD: feature fusion single shot multibox detector[J]. Computer Vision and Pattern Recognition, 2018, 36 (7): 356- 366.
9	FU C, LIU W, RANGA A, et al. DSSD: deconvolutional single shot detector[EB/OL]. [2020-07-16]. http://arxiv.org/abs/1701.06659.
10	REDMON J, DIVVALA S K, GIRSHICK R, et al. You only look once: unified, real-time object detection[C]//Proc. of the Computer Vision and Pattern Recognition, 2016: 779-788.
11	REDMON J, FARHADI A. YOLO9000: better, faster, stronger[C]//Proc. of the Computer Vision and Pattern Recognition, 2017: 6517-6525.
12	REDMON J, FARHADI A. YOLOv3: an incremental improvement[EB/OL]. [2020-07-16]. http://arxiv.org/abs/1804.02767.
13	HUANG L C, YANG Y, DENG Y F, et al. DenseBox: unifying landmark localization with end to end object detection[EB/OL]. [2020-07-16]. http://export.arxiv.org/abs/1509.04874.
14	TIAN Z, SHEN C H, CHEN H, et al. FCOS: fully convolutional one-stage object detection[EB/OL]. [2020-07-16]. http://arxiv.org/abs/1904.01355, .
15	LAW H, DENG J. CornerNet: detecting objects as paired keypoints[C]//Proc. of the European Conference on Computer Vision, 2018: 765-781.
16	李健伟, 曲长文, 彭书娟, 等. 基于卷积神经网络的SAR图像舰船目标检测[J]. 系统工程与电子技术, 2018, 40 (9): 1953- 1959.
	LI J W , QU C W , PENG S J , et al. Ship detection in SAR images based on convolutional neural network[J]. Systems Engineering and Electronics, 2018, 40 (9): 1953- 1959.
17	苏娟, 杨龙, 黄华, 等. 用于SAR图像小目标舰船检测的改进SSD算法[J]. 系统工程与电子技术, 2020, 42 (5): 1026- 1034.
	SU J , YANG L , HUANG H , et al. Improved SSD algorithm for small-scale SAR ship detection[J]. Systems Engineering and Electronics, 2020, 42 (5): 1026- 1034.
18	张晓玲, 张天文, 师君, 等. 基于深度分离卷积神经网络的高速高精度SAR舰船检测[J]. 雷达学报, 2019, 8 (6): 841- 851.
	ZHANG X L , ZHANG T W , SHI J , et al. High-speed and high-accurate SAR ship detection based on a depthwise separable convolution neural network[J]. Journal of Radars, 2019, 8 (6): 841- 851.
19	LIN T, DOLLAR P, GIRSHICK R, et al. Feature pyramid networks for object detection[C]//Proc. of the Computer Vision and Pattern Recognition, 2017: 936-944.
20	HE K, ZHANG X, REN S, et al. Deep residual learning for image recognition[C]//Proc. of the Computer Vision and Pattern Recognition, 2016: 770-778.
21	HE T, ZHANG Z, ZHANG H, et al. Bag of tricks for image classification with convolutional neural networks[C]//Proc. of the Computer Vision and Pattern Recognition, 2019: 558-567.
22	DAI J, QI H, XIONG Y, et al. Deformable convolutional networks[C]//Proc. of the International Conference on Computer Vision, 2017: 764-773.
23	ZHU X, HU H, LIN S, et al. Deformable ConvNets V2: more deformable, better results[C]//Proc. of the Computer Vision and Pattern Recognition, 2019: 9308-9316.
24	MA N, ZHANG X, ZHENG H, et al. ShuffleNet V2: practical guidelines for efficient CNN architecture design[C]//Proc. of the European Conference on Computer Vision, 2018: 122-138.
25	LI J, QU C, SHAO J. Ship detection in SAR images based on an improved faster R-CNN[C]//Proc. of the IEEE SAR in Big Data Era: Models, Methods and Applications, 2017: 1-6.
26	WANG Y , WANG C , ZHANG H , et al. A SAR dataset of ship detection for deep learning under complex backgrounds[J]. Remote Sensing, 2019, 11 (7): 765- 769. doi: 10.3390/rs11070765
27	MCMAHAN B, STREETER M J. Delay-tolerant algorithms for asynchronous distributed online learning[C]//Proc. of the Neural Information Processing Systems, 2014: 2915-2923.
28	DUCHI J , HAZAN E , SINGER Y . Adaptive subgradient methods for online learning and stochastic optimization[J]. Journal of Machine Learning Research, 2011, 12 (7): 257- 269.
29	TAN C, SUN F, KONG T, et al. A survey on deep transfer learning[C]//Proc. of the International Conference on Artificial Neural Networks, 2018: 270-279.

模型	GT	TP	FN	FP
Darknet53	267	255	12	32
Shufflenetv2	267	257	10	29
ResNet50	267	259	8	38
ResNet50-d	267	258	9	31
ResNet50-d-DCN	267	261	6	29

模型	P_d	P_MA	P_LA	Precision	mAP	大小/MB
Darknet53	95.51	4.49	11.15	88.85	93.21	234
Shufflenetv2	96.25	3.75	10.14	89.86	92.84	27
ResNet50	97.00	3.00	12.79	87.21	95.02	172
ResNet50-d	96.63	3.37	10.73	89.27	95.23	172
ResNet50-d-DCN	97.75	2.25	10.00	90.00	96.64	173

[1]	宋爽, 张悦, 张琳娜, 岑翼刚, 李浥东. 基于深度学习的轻量化目标检测算法[J]. 系统工程与电子技术, 2022, 44(9): 2716-2725.
[2]	苗添, 曾虹程, 王贺, 陈杰. 基于迭代阈值分割的星载SAR洪水区域快速提取[J]. 系统工程与电子技术, 2022, 44(9): 2760-2768.
[3]	张涛, 杨小冈, 卢瑞涛, 谢学立, 刘闯. 基于关键点的遥感图像舰船目标检测[J]. 系统工程与电子技术, 2022, 44(8): 2437-2447.
[4]	王彩云, 吴钇达, 王佳宁, 马璐, 赵焕玥. 基于改进的CNN和数据增强的SAR目标识别[J]. 系统工程与电子技术, 2022, 44(8): 2483-2487.
[5]	傅东宁, 廖桂生, 黄岩, 张邦杰, 王幸. 基于图拉普拉斯嵌入的合成孔径雷达时变窄带干扰抑制算法[J]. 系统工程与电子技术, 2022, 44(6): 1846-1853.
[6]	盖明慧, 张苏, 孙卫天, 倪育德, 杨磊. 复数兼容全变分SAR目标结构特征增强[J]. 系统工程与电子技术, 2022, 44(6): 1862-1872.
[7]	徐安林, 张毓, 周峰. 基于Beta过程的高分辨ISAR成像[J]. 系统工程与电子技术, 2022, 44(6): 1873-1879.
[8]	韦娟, 杨皇卫, 宁方立. 基于NMF与CNN联合优化的声学场景分类[J]. 系统工程与电子技术, 2022, 44(5): 1433-1438.
[9]	纪朋徽, 代大海, 邢世其, 冯德军. 密集虚假运动目标生成方法[J]. 系统工程与电子技术, 2022, 44(5): 1502-1511.
[10]	刘丰恺, 黄大荣, 郭新荣, 冯存前. 基于吕氏分布的机动目标参数化平动补偿方法[J]. 系统工程与电子技术, 2022, 44(4): 1166-1173.
[11]	陈冬, 句彦伟. 基于语义分割实现的SAR图像舰船目标检测[J]. 系统工程与电子技术, 2022, 44(4): 1195-1201.
[12]	周晓玲, 张朝霞, 鲁雅, 王倩, 王琨琨. 基于改进R-FCN的SAR图像识别[J]. 系统工程与电子技术, 2022, 44(4): 1202-1209.
[13]	方伟, 王玉, 闫文君, 林冲. 基于神经网络的符号化飞行动作识别[J]. 系统工程与电子技术, 2022, 44(3): 737-745.
[14]	孙晶明, 虞盛康, 孙俊. 基于深度学习的HRRP识别姿态敏感性分析[J]. 系统工程与电子技术, 2022, 44(3): 802-807.
[15]	杨磊, 张苏, 盖明慧, 方澄. 高分辨SAR目标成像方向性结构特征增强[J]. 系统工程与电子技术, 2022, 44(3): 808-818.