基于锚框自适应和多尺度增强的SAR舰船检测

doi:10.12305/j.issn.1001-506X.2024.04.08

摘要/Abstract

摘要：

针对目前基于深度学习的合成孔径雷达(synthetic aperture radar, SAR)舰船检测锚框尺度固定、多尺度检测性能较差的问题, 提出了一种基于锚框自适应和多尺度增强网络(adaptive anchor multi-scale enhancement network, AA-MSE-Net)的SAR舰船检测方法。首先, AA-MSE-Net引入了锚框自适应机制, 来生成适应目标形态的高质量锚框, 增强了舰船形态描述能力。其次, AA-MSE-Net提出了多尺度增强金字塔网络, 来融合增强多尺度特征, 增强了多尺度描述能力。最后, AA-MSE-Net在骨干网络中引入了可变形卷积, 来提取舰船形变特征, 进一步提高检测精度。实验证明, AA-MSE-Net在公开SAR舰船检测数据集上的平均精度高于8种对比方法。

关键词: 合成孔径雷达, 舰船检测, 自适应锚框, 尺度增强

Abstract:

Aiming at the problem that the scales of synthetic aperture radar (SAR) ship detection anchors are fixed, and the performance of multi-scale detection is poor, a SAR ship detection method based on adaptive anchors multi-scale enhancement network(AA-MSE-Net) is proposed. Firstly, AA-MSE-Net introduces adaptive anchors mechanism to generate high quality anchors that adapt to the target shape, which enhances the ability to describe the target shape. Secondly, AA-MSE-Net proposes a multi-scale enhancement feature pyramid network to fusion enhanced multi-scale features, thus enhance the multi-scale description ability of the network. Finally, AA-MSE-Net introduces deformable convolution in the backbone to extract ship deformation features and further improve detection accuracy. Experiments show that the average precision of AA-MSE-Net on the public SAR ship detection dataset is higher than that of eight comparison methods.

Key words: synthetic aperture radar (SAR), ship detection, adaptive anchor, scale enhancement

中图分类号:

TP751

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.

Zikang SHAO, Xiaoling ZHANG, Tianwen ZHANG, Tianjiao ZENG. SAR ship detection based on adaptive anchor and multi-scale enhancement[J]. Systems Engineering and Electronics, 2024, 46(4): 1204-1211.

图/表 12

图1

图2

图3

图4

图5

图6

图7

图8

图9

表1

图10

表2

参考文献 41

1	ZHANG T W , ZHANG X L , SHI J , et al. HyperLi-Net: a hyper-light deep learning network for high-accurate and high-speed ship detection from synthetic aperture radar imagery[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2020, 167, 123- 153. doi: 10.1016/j.isprsjprs.2020.05.016
2	ZHANG T W , ZHANG X L . ShipDeNet-20: an only 20 convolution layers and < 1 MB lightweight SAR ship detector[J]. IEEE Geoscience and Remote Sensing Letters, 2021, 18 (7): 1234- 1238. doi: 10.1109/LGRS.2020.2993899
3	ZHANG T W , ZHANG X L , KE X . Quad-FPN: a novel quad feature pyramid network for SAR ship detection[J]. Remote Sensing, 2021, 13 (14): 2771. doi: 10.3390/rs13142771
4	ZHANG T W , ZHANG X L . A polarization fusion network with geometric feature embedding for SAR ship classification[J]. Pattern Recognition, 2021, 123, 108365.
5	LIU T , ZHANG J F , GAO G , et al. CFAR ship detection in polarimetric synthetic aperture radar images based on whitening filter[J]. IEEE Trans.on Geoscience and Remote Sensing, 2019, 58 (1): 58- 81.
6	PELICH R , CHINI M , HOSTACHE R , et al. Large-scale automatic vessel monitoring based on dual-polarization sentinel-1 and AIS data[J]. Remote Sensing, 2019, 11 (9): 1078. doi: 10.3390/rs11091078
7	WANG C L , BI F K , ZHANG W P , et al. An intensity-space domain CFAR method for ship detection in HR SAR images[J]. IEEE Geoscience and Remote Sensing Letters, 2017, 14 (4): 529- 533. doi: 10.1109/LGRS.2017.2654450
8	ZHANG T W , ZHANG X L . Injection of traditional hand-crafted features into modern CNN-based models for SAR ship classification: what, why, where, and how[J]. Remote Sensing, 2021, 13 (11): 2091. doi: 10.3390/rs13112091
9	ZHANG T W , ZHANG X L . Squeeze-and-excitation Laplacian pyramid network with dual-polarization feature fusion for ship classification in SAR images[J]. IEEE Geoscience and Remote Sensing Letters, 2022, 19, 1- 5.
10	贾晓雅, 汪洪桥, 杨亚聃, 等. 基于YOLO框架的无锚框SAR图像舰船目标检测[J]. 系统工程与电子技术, 2022, 44 (12): 3703- 3709. doi: 10.12305/j.issn.1001-506X.2022.12.14
	JIA X Y , WANG H Q , YANG Y R , et al. Anchor free SAR image ship target detection method based on the YOLO framework[J]. Systems Engineering and Electronics, 2022, 44 (12): 3703- 3709. doi: 10.12305/j.issn.1001-506X.2022.12.14
11	REDMON J, FARHADI A. Yolov3: an incremental improvement[EB/OL]. [2022-12-05]. http://arXiv.org/abs/1804.02767.
12	GAO F , SHI W , WANG J , et al. Enhanced feature extraction for ship detection from multi-resolution and multi-scene synthetic aperture radar (SAR) images[J]. Remote Sensing, 2019, 11 (22): 2694. doi: 10.3390/rs11222694
13	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.
14	SHAO Z K , ZHANG X L , ZHANG T W , et al. RBFA-Net: a rotated balanced feature-aligned network for rotated SAR ship detection and classification[J]. Remote Sensing, 2022, 14 (14): 3345. doi: 10.3390/rs14143345
15	ZHANG T W , ZHANG X L . A full-level context squeeze-and-excitation ROI extractor for SAR ship instance segmentation[J]. IEEE Geoscience and Remote Sensing Letters, 2022, 19, 1- 5.
16	GIRSHICK R. Fast R-CNN[C]//Proc. of the IEEE International Conference on Computer Vision, 2015: 1440-1448.
17	CAI Z W , VASCONCELOS N . Cascade R-CNN: high quality object detection and instance segmentation[J]. IEEE Trans.on Pattern Analysis and Machine Intelligence, 2019, 43 (5): 1483- 1498.
18	ZHANG T W , ZHANG X L . High-speed ship detection in SAR images based on a grid convolutional neural network[J]. Remote Sensing, 2019, 11 (10): 1206. doi: 10.3390/rs11101206
19	ZHANG T W , ZHANG X L , SHI J , et al. Balance scene learning mechanism for offshore and inshore ship detection in SAR images[J]. IEEE Geoscience and Remote Sensing Letters, 2020, 19, 4004905.
20	WANG J Q, CHEN K, YANG S, et al. Region proposal by guided anchoring[C]//Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2019: 2965-2974.
21	BAO J Y , ZHANG X L , ZHANG T W , et al. ShadowDeNet: a moving target shadow detection network for video SAR[J]. Remote Sensing, 2022, 14 (2): 320. doi: 10.3390/rs14020320
22	ZHOU D F, FANG J, SONG X B, et al. IOU loss for 2D/3D object detection[C]//Proc. of the International Conference on 3D Vision, 2019: 85-94.
23	ZHANG T W , ZHANG X L . HTC+ for SAR ship instance segmentation[J]. Remote Sensing, 2022, 14 (10): 2395. doi: 10.3390/rs14102395
24	ZHANG T W , ZHANG X L , KE X , et al. LS-SSDD-v1.0: a deep learning dataset dedicated to small ship detection from large-scale Sentinel-1 SAR images[J]. Remote Sensing, 2020, 12 (18): 2997. doi: 10.3390/rs12182997
25	ZHANG T W , ZHANG X L . A mask attention interaction and scale enhancement network for SAR ship instance segmentation[J]. IEEE Geoscience and Remote Sensing Letters, 2022, 19, 1- 5.
26	PENG H , TAN X D . Improved YOLOX's anchor-free SAR image ship target detection[J]. IEEE Access, 2022, 10, 70001- 70015. doi: 10.1109/ACCESS.2022.3188387
27	张冬冬, 王春平, 付强. 基于特征增强网络的SAR图像舰船目标检测[J]. 系统工程与电子技术, 2023, 45 (4): 1032- 1039.
	ZHANG D D , WANG C P , FU Q . Ship target detection in SAR image based on feature-enhanced network[J]. Systems Engineering and Electronics, 2023, 45 (4): 1032- 1039.
28	刘万军, 高健康, 曲海成, 等. 多尺度特征增强的遥感图像舰船目标检测[J]. 国土资源遥感, 2021, 33 (3): 97- 106.
	LIU W J , GAO J K , QU H C , et al. Ship detection based on multi-scale feature enhancement of remote sensing images[J]. Remote Sensing for Natural Resources, 2021, 33 (3): 97- 106.
29	ZHANG T W , ZHANG X L , SHI J , et al. Depthwise separable convolution neural network for high-speed SAR ship detection[J]. Remote Sensing, 2019, 11 (21): 2483. doi: 10.3390/rs11212483
30	LIU S, QI L, QIN H F, et al. Path aggregation network for instance segmentation[C]//Proc. of the IEEE Conference on Computer Vision and Pattern Recognition, 2018: 8759-8768.
31	PANG J M, CHEN K, SHI J, et al. Libra R-CNN: towards balanced learning for object detection[C]//Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2019: 821-830.
32	DAI J F, QI H Z, XIONG Y W, et al. Deformable convolutional networks[C]//Proc. of the IEEE International Conference on Computer Vision, 2017: 764-773.
33	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.
34	ZHANG T W , ZHANG X L , LI J W , et al. SAR ship detection dataset (SSDD): official release and comprehensive data analysis[J]. Remote Sensing, 2021, 13 (18): 3690. doi: 10.3390/rs13183690
35	ZHANG T W , ZHANG X L , KE X , et al. HOG-ShipCLSNet: a novel deep learning network with HOG feature fusion for SAR ship classification[J]. IEEE Trans.on Geoscience and Remote Sensing, 2021, 60, 1- 22.
36	EVERINGHAM M , ESLAMI S M A , VAN G L , et al. The pascal visual object classes challenge: a retrospective[J]. International Journal of Computer Vision, 2015, 111 (1): 98- 136. doi: 10.1007/s11263-014-0733-5
37	ZHANG T W , ZHANG X L , LIU C , et al. Balance learning for ship detection from synthetic aperture radar remote sensing imagery[J]. ISPRS Journal of Photogrammetry and Remote Sensing, 2021, 182, 190- 207. doi: 10.1016/j.isprsjprs.2021.10.010
38	WEI S J , SU H , MING J , et al. Precise and robust ship detection for high-resolution SAR imagery based on HR-SDNet[J]. Remote Sensing, 2020, 12 (1): 167. doi: 10.3390/rs12010167
39	WU Y, CHEN Y P, YUAN L, et al. Rethinking classification and localization for object detection[C]//Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2020: 10186-10195.
40	ZHANG X S, WAN F, LIU C, et al. FreeAnchor: learning to match anchors for visual object detection[C]//Proc. of the 33rd International Conference on Neural Information Processing Systems, 2019: 147-155.
41	LU X, LI B Y, YUE Y X, et al. Grid RCNN[C]//Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2019: 7363-7372.

方法	Recall	Precision	AP
Faster RCNN^[16]	90.44	87.08	89.74
YOLOv3^[11]	76.10	97.18	75.68
HR-SDNet^[38]	90.99	96.49	90.82
Cascade RCNN^[17]	90.81	94.10	90.50
Libra RCNN^[31]	91.73	86.78	90.88
Double-Head RCNN^[39]	91.91	86.96	91.10
Free-Anchor^[40]	92.65	72.31	91.04
Grid RCNN^[41]	89.71	87.77	88.92
AA-MSE-Net	93.41	87.18	92.97

自适应锚框	MSE-FPN	D-Conv	Recall	Precision	AP
×	×	×	90.44	87.08	89.74
√	×	×	92.12	82.59	91.29
√	√	×	92.31	82.22	91.66
√	√	√	93.41	87.18	92.97

[1]	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.
[2]	方小宇, 黄丽佳. 基于全局位置信息和残差特征融合的SAR船舶检测算法[J]. 系统工程与电子技术, 2024, 46(3): 839-848.
[3]	张亚丽, 冯伟, 全英汇, 邢孟道. 基于多源遥感图像多级协同融合的舰船识别算法[J]. 系统工程与电子技术, 2024, 46(2): 407-418.
[4]	刘燊文, 崔兴超, 陈思伟. 结合空时上下文信息的视频SAR图像相干斑滤波[J]. 系统工程与电子技术, 2024, 46(2): 446-458.
[5]	陈洋, 肖国尧, 全英汇, 任爱锋, 别博文, 邢孟道. 基于多核DSP的星载双基FMCW SAR成像算法实现[J]. 系统工程与电子技术, 2024, 46(1): 121-129.
[6]	刘鸣谦, 徐仲秋, 陈天成, 张冰尘, 吴一戎. 基于L₁ & TV正则化的低过采样Staggered SAR成像方法[J]. 系统工程与电子技术, 2023, 45(9): 2718-2726.
[7]	王中宝, 尹奎英. 基于联合域滤波的无人机载SAR图像块效应抑制方法[J]. 系统工程与电子技术, 2023, 45(9): 2768-2776.
[8]	李海军, 孔繁程, 林云. 基于改进YOLOv5s的红外舰船检测算法[J]. 系统工程与电子技术, 2023, 45(8): 2415-2422.
[9]	寇鹏, 刘永祥, 张弛, 李玮杰, 张双辉, 霍凯. 序列ISAR像复杂结构航天器在轨姿态估计[J]. 系统工程与电子技术, 2023, 45(8): 2438-2445.
[10]	王宁, 贺鹏超, 卢景月, 刘曦. 基于DOA估计的前视多通道SAR成像方法[J]. 系统工程与电子技术, 2023, 45(8): 2471-2478.
[11]	王鹏飞, 詹珩艺, 孙洪忠. 双基前视雷达二维空变补偿频域成像方法[J]. 系统工程与电子技术, 2023, 45(7): 1990-2001.
[12]	朱瀚神, 胡文华, 郭宝锋, 焦丽婷, 朱晓秀, 朱常安. 双基地ISAR稀疏孔径机动目标MTRC补偿成像算法[J]. 系统工程与电子技术, 2023, 45(7): 2022-2030.
[13]	汪俊澎, 邢世其, 李永祯, 黄大通, 宋少秋. 基于时频交叉乘积的调频连续波SAR干扰方法研究[J]. 系统工程与电子技术, 2023, 45(6): 1651-1657.
[14]	杨磊, 夏亚波, 廖仙华, 毛欣瑶, 窦宇宸, 杨桓. 双层稀疏贝叶斯学习ISAR超分辨成像算法[J]. 系统工程与电子技术, 2023, 45(5): 1371-1379.
[15]	吴其华, 赵锋, 王俊杰, 徐志明, 肖顺平. 基于编码相位调制的ISAR图像特征控制方法[J]. 系统工程与电子技术, 2023, 45(4): 1000-1007.