基于全局特征融合的SAR图像旋转舰船目标检测算法

doi:10.12305/j.issn.1001-506X.2024.12.13

摘要/Abstract

摘要：

针对常规模型在合成孔径雷达(synthetic aperture radar, SAR)图像中对近岸旋转舰船目标检测效果不佳的问题, 提出一种基于全局特征融合的SAR图像旋转舰船目标检测方法。首先, 通过全局注意力特征金字塔网络融合不同层级特征, 缩短了底层特征向顶层特征的传递路径。其次, 在图像块融合阶段加入位置编码，以减少降采样导致的定位信息损失。最后, 采用旋转特征对齐网络生成高质量的锚点和旋转对齐特征, 用于分类和坐标回归。所提方法在SAR舰船斜框检测数据集(rotated ship detection dataset in SAR images, RSDD-SAR) 上旋转交并比为0.5时的平均检测精度达到了0.894 8, 对近岸和离岸舰船都有着较好的检测性能。

关键词: 合成孔径雷达, 舰船检测, 旋转目标检测, 神经网络, 特征融合

Abstract:

Conventional models are not effective in detecting inshore rotated ship targets in synthetic aperture radar (SAR) images. To solve this problem, a method of detecting rotated ship targets in SAR images based on global feature fusion is proposed. Firstly, the global attention feature pyramid network is used to fuse features of different levels, which shortens the transmission path from the bottom feature to the top feature. Secondly, positional embedding is added in the image block fusion stage to reduce the loss of location information caused by down-sampling. Finally, the rotated feature alignment network is used to generate high-quality anchor points and rotated alignment features for classification and coordinate regression. The proposed method achieves an average detection precision of 0.894 8 on the rotated ship detection dataset in SAR images (RSDD-SAR) dataset when the rotated intersection over union (IoU) is 0.5. The proposed method has good detection performance for both inshore and offshore ships.

Key words: synthetic aperture radar (SAR), ship detection, rotated target detection, neural network, feature fusion

中图分类号:

TP751.1

薛峰涛, 孙天宇, 杨益民, 杨健. 基于全局特征融合的SAR图像旋转舰船目标检测算法[J]. 系统工程与电子技术, 2024, 46(12): 4044-4053.

Fengtao XUE, Tianyu SUN, Yimin YANG, Jian YANG. Rotated ship target detection algorithm in SAR images based on global feature fusion[J]. Systems Engineering and Electronics, 2024, 46(12): 4044-4053.

图/表 17

图1

图2

图3

图4

图5

图6

图7

图8

图9

表1

表2

图10

表3

图11

图12

图13

表4

参考文献 30

1	邓云凯, 赵凤军, 王宇. 星载SAR技术的发展趋势及应用浅析[J]. 雷达学报, 2012, 1 (1): 1- 10.
	DENG Y K , ZHAO F J , WANG Y . Brief analysis on the development and application of spaceborne SAR[J]. Journal of Radars, 2012, 1 (1): 1- 10.
2	邓云凯, 禹卫东, 张衡, 等. 未来星载SAR技术发展趋势[J]. 雷达学报, 2020, 9 (1): 1- 33.
	DENG Y K , YU W D , ZHANG H , et al. Forthcoming spaceborne SAR development[J]. Journal of Radars, 2020, 9 (1): 1- 33.
3	GIOVANNI M P , MARCO M , STEFANO T , et al. Very high resolution automotive SAR imaging from burst data[J]. Remote Sensing, 2023, 15 (3): 845- 864. doi: 10.3390/rs15030845
4	李震宇. 机动平台SAR大斜视成像算法研究[D]. 西安: 西安电子科技大学, 2017.
	LI Z Y. Study on high squint imaging algorithms for SAR mounted on maneuvering platforms[D]. Xi'an: Xidian University, 2017.
5	YAN G X , CHEN Z H , WANG Y , et al. LssDet: a lightweight deep learning detector for SAR ship detection in high-resolution SAR images[J]. Remote Sensing, 2022, 14 (20): 5148- 5166. doi: 10.3390/rs14205148
6	ZHAO K , LU R T , WANG S Y , et al. ST-YOLOA: a Swin-Transformer-based YOLO model with an attention mechanism for SAR ship detection under complex background[J]. Frontiers in Neurorobotics, 2023, 17, 1170163- 1170177. doi: 10.3389/fnbot.2023.1170163
7	李焘. 基于SAR图像的舰船目标检测方法研究[D]. 西安: 西安电子科技大学, 2019.
	LI T. Study on ship detection methods for SAR images[D]. Xi'an: Xidian University, 2019.
8	LIU T , YANG Z Y , YANG J , et al. CFAR ship detection methods using compact polarimetric SAR in a K-Wishart distri- bution[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2019, 12 (10): 3737- 3745. doi: 10.1109/JSTARS.2019.2923009
9	LI J W , CHEN J , CHENG P , et al. A survey on deep-learning-based real-time SAR ship detection[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2023, 16, 3218- 3247. doi: 10.1109/JSTARS.2023.3244616
10	LIN T Y, GOYAL P, GIRSHICK R, et al. Focal loss for dense object detection[C]//Proc. of the IEEE International Conference on Computer Vision, 2017: 2980-2988.
11	TIAN Z, SHEN C H, CHEN H, et al. FCOS: fully convolutional one-stage object detection[C]//Proc. of the IEEE/CVF International Conference on Computer Vision, 2019: 9627-9636.
12	CHANG Y L , ANAGAW A , CHANG L , et al. Ship detection based on YOLOv2 for SAR imagery[J]. Remote Sensing, 2019, 11 (7): 786- 799. doi: 10.3390/rs11070786
13	LI Y D , ZHANG S S , WANG W Q . A lightweight faster R-CNN for ship detection in SAR images[J]. IEEE Geoscience and Remote Sensing Letters, 2020, 19, 4006105.
14	MAO Y X, LI X J, LI Z L, et al. An anchor-free SAR ship detector with only 1.17 m parameters[C]//Proc. of the International Conference on Aviation Safety and Information Technology, 2020: 182-186.
15	SUN Z Z , DAI M C , LENG X G , et al. An anchor-free detection method for ship targets in high-resolution SAR images[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2021, 14, 7799- 7816. doi: 10.1109/JSTARS.2021.3099483
16	YU J M , ZHOU G Y , ZHOU S B , et al. A fast and lightweight detection network for multi-scale SAR ship detection under complex back-grounds[J]. Remote Sensing, 2022, 14 (1): 31- 54.
17	徐从安, 苏航, 李健伟, 等. RSDD-SAR: SAR舰船斜框检测数据集[J]. 雷达学报, 2022, 11 (4): 581- 599.
	XU C A , SU H , LI J W , et al. RSDD-SAR: rotated ship detection dataset in SAR images[J]. Journal of Radars, 2022, 11 (4): 581- 599.
18	ZHOU Y , JIANG X , XU G Z , et al. PVT-SAR: an arbitrarily oriented SAR ship detector with pyramid vision transformer[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2022, 16, 291- 305.
19	ZHOU K X , ZHANG M , ZHAO H H , et al. Arbitrary-oriented ellipse detector for ship detection in remote sensing images[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2023, 16, 7151- 7162. doi: 10.1109/JSTARS.2023.3267240
20	GE J Y , TANG Y P , GUO K T , et al. KeyShip: towards high-precision oriented SAR ship detection using key points[J]. Remote Sensing, 2023, 15 (8): 2035- 2060. doi: 10.3390/rs15082035
21	ZHANG Y P , LU D D , QIU X L , et al. Scattering-point-guided RPN for oriented ship detection in SAR images[J]. Remote Sensing, 2023, 15 (5): 1411- 1432. doi: 10.3390/rs15051411
22	HAN J M , DING J , LI J , et al. Align deep features for oriented object detection[J]. IEEE Trans.on Geoscience and Remote Sensing, 2021, 60, 5602511.
23	HE K M, ZHANG X Y, REN S Q, et al. Deep residual learning for image recognition[C]//Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2016: 770-778.
24	LIU Z, LIN Y T, CAO Y, et al. Swin-Transformer: hierarchical vision transformer using shifted windows[C]//Proc. of the IEEE/CVF International Conference on Computer Vision, 2021: 10012-10022.
25	DOSOVITSKIY A, BEYER L, KOLESNI-KOV A, et al. An image is worth 16×16 words: transformers for image recognition at scale[EB/OL]. [2023-06-05]. http://arxiv.org/abs2010.11929.
26	LIN T Y, DOLLAR P, GIRSHICK R, et al. Feature pyramid networks for object detection[C]// Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2017: 2117-2125.
27	YANG S, PEI Z Q, ZHOU F, et al. Rotated faster R-CNN for oriented object detection in aerial images[C]//Proc. of the 3rd International Conference on Robot Systems and Applications, 2020: 35-39.
28	LIN T Y, GOYAL P, GIRSHICK R, et al. Focal loss for dense object detection[C]//Proc. of the IEEE International Conference on Computer Vision, 2017: 2980-2988.
29	XIE X X, CHENG G, WANG J B, et al. Oriented R-CNN for object detection[C]//Proc. of the IEEE/CVF International Conference on Computer Vision, 2021: 3520-3529.
30	YANG X, YAN J C, FENG Z M, et al. R3Det: refined single-stage detector with feature refinement for rotating object[EB/OL]. [2023-05-10]. https://arxiv.org/pdf/1908.05612v4.pdf.

混淆矩阵		真实类别
混淆矩阵		正样本	负样本
预测类别	正样本	TP	FP
预测类别	负样本	FN	TN

IoU	P	R	AP
0.50	0.886	0.896	0.895
0.60	0.868	0.799	0.842
0.70	0.777	0.711	0.734
0.80	0.612	0.404	0.449
0.90	0.391	0.255	0.231

方法	骨干网络	AP₅₀	运算次数	每秒帧数	参数量大小
Rotated Faster RCNN	ResNet50	0.831 5	63.26	62.4	41.14
Rotated RetinaNet	ResNet50	0.861 4	52.39	75.4	36.13
Oriented RCNN	ResNet50	0.883 7	63.28	62.4	41.13
S2ANet	ResNet50	0.873 2	49.05	80.5	38.76
Rotated FCOS	ResNet50	0.859 5	51.55	76.6	32.12
R3Det	ResNet50	0.812 2	82.17	48.1	41.81
本文方法	Swin-Tiny	0.894 8	79.06	49.9	58.65

骨干网络	特征融合	位置编码	AP
ResNet50	-	-	0.869 2
Swin-Tiny	-	-	0.882 5
Swin-Tiny	√	-	0.891 4
Swin-Tiny	√	√	0.894 8

[1]	张天骐, 杨宗方, 邹涵, 马焜然. 基于编码矩阵估计的极化码参数盲识别算法[J]. 系统工程与电子技术, 2024, 46(9): 3221-3230.
[2]	王磊, 张劲, 叶秋炫. LDACS系统基于循环谱和残差神经网络的频谱感知方法[J]. 系统工程与电子技术, 2024, 46(9): 3231-3238.
[3]	颜上取, 付耀文, 张文鹏, 杨威, 余若峰, 张法桐. 视频合成孔径雷达技术发展现状综述[J]. 系统工程与电子技术, 2024, 46(8): 2650-2666.
[4]	王进, 冷祥光, 孙忠镇, 马晓杰, 杨阳, 计科峰. 复杂运动舰船目标SAR成像空/时变散焦特性研究[J]. 系统工程与电子技术, 2024, 46(7): 2237-2255.
[5]	张瑞斌, 朱梦韬, 李云杰. 对调制识别网络隐身的雷达发射信号生成方法[J]. 系统工程与电子技术, 2024, 46(7): 2256-2268.
[6]	齐兵, 程建华, 赵砚驰, 汪籽粒. 基于微形变分析的电容式MEMS加速度计温漂误差精密估计方法[J]. 系统工程与电子技术, 2024, 46(7): 2437-2445.
[7]	邢世其, 纪朋徽, 代大海, 冯德军. 方位向调制干扰对高分宽幅多通道SAR的影响[J]. 系统工程与电子技术, 2024, 46(6): 1946-1956.
[8]	吴巍屹, 贾云献, 姜相争, 史宪铭, 刘洁, 刘彬, 董恩志, 朱曦. 面向阶段任务的携行器材品种确定方法[J]. 系统工程与电子技术, 2024, 46(6): 2054-2064.
[9]	曾顶, 殷君君, 杨健. 基于融合距离的极化SAR图像非局部均值滤波[J]. 系统工程与电子技术, 2024, 46(5): 1493-1502.
[10]	庆豪, 方志耕, 王育红, 邱玺睿. 基于灰色-神经网络的民机需求组合预测[J]. 系统工程与电子技术, 2024, 46(5): 1665-1672.
[11]	杨宗方, 张天骐, 马焜然, 邹涵. 基于深层神经网络的信道编码类型盲识别[J]. 系统工程与电子技术, 2024, 46(5): 1820-1829.
[12]	邵子康, 张晓玲, 张天文, 曾天娇. 基于锚框自适应和多尺度增强的SAR舰船检测[J]. 系统工程与电子技术, 2024, 46(4): 1204-1211.
[13]	何通, 卢青, 周军, 郭宗易. 带有神经网络干扰观测器的视线角约束制导[J]. 系统工程与电子技术, 2024, 46(4): 1372-1382.
[14]	周红进, 宋辉, 范文良, 王苏, 谷东亮. 基于贝叶斯神经网络的船用惯导定位修正方法[J]. 系统工程与电子技术, 2024, 46(4): 1393-1400.
[15]	张天文, 张晓玲, 邵子康, 曾天娇. 基于掩模注意型交互的SAR舰船实例分割[J]. 系统工程与电子技术, 2024, 46(3): 831-838.