基于证据理论的多源图像目标检测方法

doi:10.12305/j.issn.1001-506X.2026.07.01

摘要/Abstract

摘要：

针对多源图像舰船目标检测中存在的样本依赖性强、异时相数据融合效果差的问题，提出一种基于D-S （Dempster-Shafer）证据理论的自适应决策融合方法。不同于传统先融合后检测的结构，该方法构建了差异化基本概率指派模型，通过将异时相样本融合产生的信息干扰转化为互补优势，有效解决了多源融合检测对同时相样本依赖性强的问题，实现多源图像的有效决策融合。结果表明，所提方法的平均精度较单源光学图像和合成孔径雷达图像检测分别提升0.7%和2.92%。对比传统融合检测算法，所提方法缩减了运算开支，显著提高了检测精度。

关键词: D-S （Dempster-Shafer）证据理论, 样本依赖, 多源融合, 目标检测

Abstract:

To address the issues of high sample dependence and poor multi-temporal data fusion effectiveness in multi-source image ship target detection, this paper proposes an adaptive decision fusion method based on D-S （Dempster-Shafer） evidence theory. Departing from the traditional fuse-then-detect framework, this method constructs a differentiated basic probability assignment model. It transforms the information interference generated by fusing samples from different time periods into complementary benefits, effectively overcoming the dependence on simultaneous-phase samples in multi-source fusion detection and achieving effective decision fusion of multi-source images. Experimental results show that the proposed method improves average precision （AP） by 0.7% and 2.92% respectively compared to single-source optical and synthetic aperture radar （SAR） image detection. Compared with conventional fusion detection algorithms, it reduces computational overhead while significantly enhancing detection accuracy.

Key words: D-S （Dempster-Shafer） evidence theory, sample dependency, multi-source fusion, target detection

中图分类号:

TP 391.41

姜娜, 刘霖, 陈世龙. 基于证据理论的多源图像目标检测方法[J]. 系统工程与电子技术, 2026, 48(7): 2133-2143.

Na JIANG, Lin LIU, Shilong CHEN. Multi-source image target detection method based on evidence theory[J]. Systems Engineering and Electronics, 2026, 48(7): 2133-2143.

图/表 13

图1

图2

图3

图4

图5

图6

图7

图8

图9

表1

图10

表2

表3

参考文献 34

1	SONG H S, WANG L, MI L, et al. DCAPT-Net: a multi-scale ship detection algorithm for SAR images[C]//Proc. of the 10th International Conference on Big Data and Information Analytics, 2024: 441−446.
2	ASHIQUZZAMAN A, OH S, LEE D, et al. Compact deeplearning convolutional neural network based hand gesture classifier application for smart mobile edge computing[C]//Proc. of the International Conference on Artificial Intelligence in Information and Communication, 2020: 119−123.
3	XU S J, JI Y J, WANG G C, et al. GFSPP-YOLO: a light YOLO model based on group fast spatial pyramid pooling[C]//Proc. of the IEEE 11th International Conference on Information, Communication and Networks, 2023: 733−738.
4	WANG H, GAO P. Survey of small object detection methods based on deep learning[C]//Proc. of the 9th International Conference on Intelligent Informatics and Biomedical Sciences, 2024: 221−224.
5	CHEN S L, LIU L, WANG X B, et al. Research on SAR active anti-jamming imaging based on joint random agility of inter-pulse multi-parameters in the presence of active deception[J]. Remote Sensing, 2024, 16 (17): 3303. doi: 10.3390/rs16173303
6	LIU S Z, LI K P, MI J H, et al. Multi-source navigation sensor backend information fusion technology based on D-S evidence theory[C]//Proc. of the 9th International Symposium on System Security, Safety, and Reliability, 2023: 197−202.
7	赵铖鑫. 联合SAR与光学图像的目标检测与识别技术[D]. 成都: 电子科技大学, 2023.
	ZHAO C X. Joint SAR and optical image target detection and recognition technology[D]. Chengdu: University of Electronic Science and Technology of China, 2023.
8	ZHONG Y T, ZHANG S Y, LIU Z Y, et al. Unsupervised fusion of misaligned PAT and MRI images via mutually reinforcing cross-modality image generation and registration[J]. IEEE Trans. on Medical Imaging, 2024, 43 (5): 1702- 1714. doi: 10.1109/TMI.2023.3347511
9	陈稳. 基于光学和SAR遥感影像融合的典型目标检测识别研究[D]. 哈尔滨: 哈尔滨工业大学2019.
	CHEN W. Research on typical target detection and recognition based on optical and SAR remote sensing image fusion[D]. Harbin: Harbin Institute of Technology, 2019.
10	CHEN J L, XU X Q, ZHANG J C, et al. Ship target detection algorithm based on decision-level fusion of visible and SAR images[J]. IEEE Journal on Miniaturization for Air and Space Systems, 2023, 4 (3): 242- 249. doi: 10.1109/JMASS.2023.3269434
11	王梓衡, 沈继锋, 左欣, 等. 基于特征级与决策级融合的农作物叶片病害识别[J]. 江苏大学学报（自然科学版）, 2024, 45 (3): 286- 294.
	WANG Z H, SHEN J F, ZUO X, et al. Crop leaf disease recognition based on feature level and decision level fusion[J]. Journal of Jiangsu University （Natural Science Edition）, 2024, 45 (3): 286- 294.
12	周策, 赵秋博, 王明杰, 等. 基于证据理论的可见光和红外融合检测算法[J]. 无线电工程, 2025, 55 (3): 484- 492.
	ZHOU C, ZHAO Q B, WANG M J, et al. Visible light and infrared fusion detection algorithm based on evidence theory[J]. Radio Engineering, 2025, 55 (3): 484- 492.
13	WANG Z Y, WANG H Q, WANG Y. Pyramid attention network for medical image registration[C]//Proc. of the IEEE International Symposium on Biomedical Imaging, 2024.
14	WU Q, MA Y N, WANG Y, et al. Comparison of pixel-level and feature-level image fusion networks for slow-moving landslide detection[C]//Proc. of the IGARSS IEEE International Geoscience and Remote Sensing Symposium, 2024: 3963−3966.
15	LIU L, SUN N, WANG J C. Spearman-coefficient-based improved D-S evidence theory algorithm considering the initial importance of evidence body and its application on multi-source information fusion[C]//Proc. of the 3rd International Conference on Computer Science, Electronic Information Engineering and Intelligent Control Technology, 2023: 312−318.
16	BECK J, KEMETER L M, DÜRRBECK K, et al. Toward integrating ChatGPT into satellite image annotation workflows: a comparison of label quality and costs of human and automated annotators[J]. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2025, 18, 4366- 4381. doi: 10.1109/JSTARS.2025.3528192
17	HU Y K, TANG Y S, BU S M, et al. Hungarian algorithm integrated with iterative method for multi-UAV penetration game strategy[C]//Proc. of the IEEE International Conference on Unmanned Systems, 2024: 1423−1428.
18	BAKIRCI M, BAYRAKTAR I. The cutting-edge YOLO11 for advanced aircraft detection in synthetic aperture radar (SAR) imagery[C]//Proc. of the 8th International Symposium on Innovative Approaches in Smart Technologies, 2024.
19	DANG Z C, WANG X S. FD-YOLO11: a feature-enhanced deep learning model for steel surface defect detection[J]. IEEE Access, 2025, 13, 63981- 63993. doi: 10.1109/ACCESS.2025.3559733
20	ZHANG B N, CHEN H, CHEN J L, et al. Design and implementation of anti-peeping screen algorithm based on soft-NMS YOLOv7[C]//Proc. of the 7th International Conference on Pattern Recognition and Artificial Intelligence, 2024: 71−76.
21	YUE D S, SHI Y C, CHEN X, et al. Wellhead assignment model and improved hungarian algorithm for large scale cluster well group[C]//Proc. of the 2nd International Conference on Big Data & Artificial Intelligence & Software Engineering, 2021: 296−299.
22	JANOS L L, EDWIN S T, JOSE J A C. Efficient vehicle counting algorithm using gaussian mixing models and IOU based tracker for LPWAN based intelligent traffic management systems[C]// Proc. of the IEEE 14th International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management, 2022.
23	CHEN C J, LUO B C, YIN K X. Research on bearing fault diagnosis based on DS evidence theory and convolutional neural network[C]//Proc. of the 5th International Seminar on Artificial Intelligence, Networking and Information Technology, 2024: 2360−2364.
24	关爱科, 杨杰. 基于SVM-DS融合算法的消防员效能评估方法[J]. 消防科学与技术, 2024, 43 (12): 1772- 1777.
	GUAN A K, YANG J. Firefighter efficiency evaluation method based on SVM-DS fusion algorithm[J]. Fire Science and Technology, 2024, 43 (12): 1772- 1777.
25	HAI B, YAO H L. An approach to information fusion and group deecision making based on choquet integral operators for interval-valued intuitionistic fuzzy data[C]//Proc. of the International Conference on Machine Learning and Cybernetics, 2018: 109−114.
26	SURATHONG S, MAISEN C, PIYAWONGWISAL P. Modified fuzzy dempster-shafer theory for decision fusion[C]//Proc. of the 13th International Conference on Information Technology and Electrical Engineering, 2021: 244−248.
27	WANG H L, WANG W S, HE J Y. Cafeteria self-service billing system based on YOLOv11 and HSV threshold segmentation algorithm[C]//Proc. of the 5th International Conference on Consumer Electronics and Computer Engineering, 2025: 306−312.
28	GIURGI D V, DEZERT J, JOSSO L, et al. Fusion of semantic segmentation models for vehicle perception tasks[C]//Proc. of the 27th International Conference on Information Fusion, 2024.
29	SUN L F, ZHANG Y Y, FU Z M, et al. An approach to multi-sensor decision fusion based on the improved jousselme evidence distance[C]//Proc. of the International Conference on Control, Automation and Information Sciences, 2018: 189−193.
30	SHERMEYER J, HOGAN D, BROWN J, et al. SpaceNet 6: multi-sensor all weather mapping dataset[C]//Proc. of the IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2020: 768−777.
31	CHEN Y S, JIA Z Y, HIROTA K R, et al. A multimodal emotion perception model based on context-aware decision-level fusion[C]//Proc. of the 41st Chinese Control Conference, 2022: 7332−7337.
32	PRABHAKAR K R, SRIKAR V S, BABU R V. DeepFuse: a deep unsupervised approach for exposure fusion with extreme exposure image pairs[C]//Proc. of the IEEE International Conference on Computer Vision, 2017: 4724−4732.
33	FU Y, WU X J. A dual-branch network for infrared and visible image fusion[EB/OL]. [2025-04-03]. http://arxiv.org/abs/2101.09643.
34	FU Y, WU X J, KITTLER J. Deep decomposition network for image processing: a case study for visible and infrared image fusion[EB/OL]. [2025-04-03]. http://arxiv.org/abs/2102.10526.

方法	数据集	AP/%	R/%	F1/%
YOLO11-obb	PSRGB	94.93	95.54	90.36
Rotate Faster R-CNN	PSRGB	84.12	88.85	84.16
ROI-Transformer	PSRGB	91.30	93.31	91.28
本文方法	PSRGB	95.63	96.50	95.73
YOLO11-obb	SAR	74.68	86.82	84.85
Rotate Faster R-CNN	SAR	42.44	56.59	45.77
ROI-Transformer	SAR	52.60	62.40	60.64
本文方法	SAR	77.60	87.21	86.71

方法	AP/%	R/%	F1/%
Deepfuse+YOLO11	72.09	85.54	84.22
Dualfuse+YOLO11	71.10	84.34	82.64
Deepdecfuse+YOLO11	78.83	89.40	85.09
WLS+YOLO11	79.69	91.57	86.46
本文方法	87.27	89.51	89.12

方法	FLOPs/G	总参数量/×10⁶	GPU占用率/%	时间/s
Deepfuse+YOLO11	19.43	2.67	11	5.34
Dualfuse+YOLO11	14.87	2.93	29	4.73
Deepdecfuse+YOLO11	28.30	2.95	48	1.41
WLS+YOLO11	6.60	2.65	4	7.88
本文方法	13.20	5.30	9	1.94

[1]	赵巍山, 尤思洋, 黄丽佳, 周光尧. 基于散射拓扑增强的双流SAR飞机检测网络[J]. 系统工程与电子技术, 2026, 48(6): 1848-1858.
[2]	张杰, 常天庆, 王晓卫, 郝文龙, 汤鑫. 基于可见光与红外特征融合的轻量化目标检测方法[J]. 系统工程与电子技术, 2026, 48(5): 1481-1491.
[3]	王新, 周生华, 张新勋, 南静怡, 孙世坤. 全极化MIMO雷达主瓣压制干扰抑制方法[J]. 系统工程与电子技术, 2026, 48(5): 1539-1550.
[4]	刘硕, 周宇, 张哲昊, 尚嵩, 程松. 时域多脉冲相关与聚类分析的抗同频干扰技术[J]. 系统工程与电子技术, 2026, 48(4): 1186-1194.
[5]	鲁明雨, 孟飞, 叶春茂, 李璋峰, 赵庆媛. 一种先验知识辅助的扩展目标检测方法[J]. 系统工程与电子技术, 2026, 48(4): 1209-1217.
[6]	程鲲, 雷洪涛, 吕志轩. 改进YOLOv8的轻量级无人机跟踪方法[J]. 系统工程与电子技术, 2026, 48(3): 737-750.
[7]	王威, 杨勇, 韩静雯. 基于相关性的雷达海面弱目标检测技术[J]. 系统工程与电子技术, 2026, 48(3): 787-794.
[8]	缪燕子, 赵志非, 吴巍. 基于代价敏感学习DBN-XGBoost的海面小目标检测方法研究[J]. 系统工程与电子技术, 2026, 48(2): 402-409.
[9]	陈凯, 赵永波, 刘仍莉, 邓海涛, 孙龙. 机载长合成孔径时间海面运动舰船高分辨SAR成像算法[J]. 系统工程与电子技术, 2026, 48(2): 456-465.
[10]	丁昊, 韦继丰, 董云龙, 曹政, 于恒力. 高海况下应用先验信息的海上小目标检测方法[J]. 系统工程与电子技术, 2026, 48(1): 44-55.
[11]	蓝舒尧, 李宇, 张春华, 迟骋, 陈春辉. 垂直阵分集束控影区探测信号发射策略[J]. 系统工程与电子技术, 2025, 47(8): 2454-2462.
[12]	于营, 王春平, 徐金辉, 吕述杭, 付强, 陈明. PE-Net：一种优化剪枝的实时山体滑坡检测网络[J]. 系统工程与电子技术, 2025, 47(8): 2475-2485.
[13]	罗颖聪, 张磊, 魏少鹏, 孟智超. 联合失配滤波器的近区低旁瓣混沌波形设计算法[J]. 系统工程与电子技术, 2025, 47(8): 2511-2518.
[14]	付卫红, 彭文洪, 刘乃安. 混合注意力优化的SAR图像小目标检测方法[J]. 系统工程与电子技术, 2025, 47(8): 2519-2526.
[15]	倪康, 贾文杰, 邹旻瑞, 郑志忠. 基于动态聚合网络的SAR目标检测[J]. 系统工程与电子技术, 2025, 47(8): 2527-2539.