Journal of Engineering Research
Journal of Engineering Research. 2026; 5: (3) ; 10.12208/j.jer.20260035 .
总浏览量: 570
中国电子科学研究院 北京
*通讯作者: 李奇峰,单位:中国电子科学研究院 北京; ;
针对红外图像中弱小目标像素占比低、特征稀疏、背景干扰严重等问题,提出一种基于多感受野特征融合的红外弱小目标检测算法。该算法以YOLOv5s为基础检测框架,通过设计多分支感受野扩展模块,在颈部网络中构建多尺度上下文信息融合机制,增强了网络对弱小目标的特征表达能力;同时引入自适应特征加权策略,有效抑制复杂背景噪声干扰。在公开红外弱小目标数据集上的实验结果表明,改进后算法相较于基线YOLOv5s,在检测精度、召回率及定位准确性等指标上均有显著提升,验证了多感受野融合策略对红外弱小目标检测任务的有效性。
To address the challenges of extremely low pixel occupancy, sparse features, and severe background interference in infrared dim and small target detection, this paper proposes an infrared dim and small target detection algorithm based on multi-receptive field feature fusion. The algorithm adopts YOLOv5 as the baseline detection framework and constructs a multi-scale contextual information fusion mechanism within neck networks by designing a multi-branch receptive field expansion module, thereby enhancing the network's feature representation capability for dim and small targets. Meanwhile, an adaptive feature weighting strategy is introduced to effectively suppress interference from complex background clutter. Experimental results on public infrared dim and small target datasets demonstrate that, compared with the baseline YOLOv5, the improved algorithm achieves significant improvements in detection precision, recall, and localization accuracy, validating the effectiveness of the multi-receptive field fusion strategy for infrared dim and small target detection tasks.
[1] 张晶.临近空间高超声速飞行器发展及防御策略研究[J].飞航导弹,2015(4):25-29.
[2] 王鹏飞,王光明,蒋坤,等.临近空间高超声速飞行器发展及关键技术研究 [J].飞航导弹,2019 (8):22-28.
[3] 李高飞,冯占林,陆泽健.单枚弹道导弹目标的威胁度评估方法研究[J].中国电子科学研究院学报,2021,16(1):46-50+61.
[4] 许有玺, 赫健, 陈雪波. 基于梯度模糊TOPSIS导弹目标的威胁评估[J]. 现代防御技术, 2024, 52(6): 33-40.
[5] 徐浩, 邢清华. 基于证据理论的助推段弹道导弹目标威胁评估[J]. 军事运筹与系统工程, 2016, 30(2): 9-14.
[6] Qiu S, Wang J, Zhang X, et al. Multi target threat assessment method based on improved dual variable weight[C]//Advanced Manufacturing and Automation X. Springer, 2021: 475-485.
[7] Zhao J, Guo Q, Teng L, et al. Application of Deep neural Network in Air Target Threat Assessment[C]//2019 IEEE International Conference on Computational Science and Engineering. IEEE, 2019.
[8] 谢常达,范玺斌,王雪峰. 基于层次分析法对舰队防空威胁度研究[J]. 舰船电子工程,2022,v.42;No.331(01):23-27+32.
[9] 汪伟,顾竹鑫,李海波. 基于层次分析法的低空慢速小目标威胁评估方法[J]. 信息化研究,2019,v.45;No.285(04): 20-24.
[10] 赵永,李为民,刘彬,崔超. 基于改进灰色关联法的高超声速目标威胁评估模型[J]. 探测与控制学报,2014,v.36; No.166(05):80-85.
[11] 杨璐,刘付显,张涛,朱丰. 基于组合赋权TOPSIS法的舰艇编队空中目标威胁评估模型[J]. 电光与控制,2019, v.26;No.254(08):6-11.
[12] 李丹一. 基于贝叶斯网络的水下威胁评估方法研究[D].浙江大学,2021.
[13] 李继广,布国亮,董彦非,袁盼盼,樊佳乐,刘天栋. 基于改进模糊层次分析法通信无人机机场选址[J]. 兵工自动化,2023,v.42;No.301(11):83-87+96.
[14] 周华任,马亚平,郭杰,等.基于五力的武器装备作战能力评估模型[J]. 火力与指挥控制,2011,36(2):11-14.
[15] 李京峰,项华春,严雅榕,李正欣. 基于离差最大化的组合赋权评价方法及其应用[J]. 火力与指挥控制,2020,v.45; No.303(06):18-23+29.