Infrared Weak Target Detection Method Based on Sparse Attention
-
摘要:
针对复杂背景下红外弱小目标像素占比少，细节纹理特征匮乏导致特征提取困难、检测率低、虚警率高的问题，提出一种基于稀疏注意力和多尺度特征融合的红外弱小目标检测网络。该网络利用Resnest的分割注意力提取不同尺度特征，引入Biformer注意力模块学习目标与背景之间的远程关系，采用融合模块将高、低层特征进行融合，经过Head模块输出检测结果二值图。实验结果表明，本文方法在IoU和Fmeasure这两项指标中均取得最优，与DNANet方法相比，所提方法的交并比（IoU）提高3.9%、Fmeasure提高5.6%；与ABCNet方法相比，所提方法的IoU提高5.8%、Fmeasure提高10%；并且在不同复杂背景下均可有效检测出红外弱小目标，体现良好的鲁棒性和适应性，可以有效应用于复杂背景中的红外弱小目标检测。
Abstract:In this study, a novel weak infrared small target detection network based on sparse attention and multiscale feature fusion is proposed to address the challenges of low pixel occupancy and limited texture features for weak infrared small targets within complex backgrounds, leading to difficulties in feature extraction, low detection rates, and high false alarm rates. The network utilizes the segmentation attention of ResNest to extract features at different scales. A BiFormer attention module is introduced to learn the distant relationships between targets and backgrounds. Furthermore, a fusion module is employed to merge both high- and low-level features, with the final detection results represented as a binary image through a head module. The experimental results demonstrate that the proposed method achieves the best performance in terms of both Intersection over Union (IoU) and F-measure. Compared with the dense nested attention network (DNANet), the proposed method improved the IoU by 3.9% and F-measure by 5.6%. Compared with the attentive bilateral contextual network (ABCNet), the proposed method improved the IoU by 5.8% and F-measure by 10%. Moreover, the proposed approach exhibited robustness and adaptability in effectively detecting small weak infrared targets in diverse, complex backgrounds. This method is applicable to weak infrared small-target detection in complex backgrounds, exhibiting superior performance.
-
Keywords:
- infrared weak target detection /
- sparse attention /
- feature fusion /
- robustness
-
-
![]()
图 1 本文的红外弱小目标检测网络结构
Figure 1. The structure diagram of the infrared weak target detection network in this paper
![]()
图 4 不同方法对各类复杂背景干扰的红外弱小目标检测结果对比
Figure 4. Comparison of infrared weak and small target detection results by different methods against various complex background interferences
![]()
图 5 本文方法与对比方法检测结果的ROC曲线
Figure 5. ROC curves comparison between the proposed method and comparative methods for detection results
表 1 不同方法对复杂背景下红外弱小目标检测结果的指标对比
Table 1 Comparison of metrics for infrared weak small target detection results under complex backgrounds using different methods
SCRG BSF IoU Pr Re Fmeasure FC3Net 66.8 26.2 0.612 0.739 0.754 0.727 AGPCNet 77.6 28.1 0.628 0.695 0.694 0.659 ACM 74.1 34.9 0.586 0.710 0.754 0.714 LSPM 89.7 25.1 0.543 0.822 0.607 0.682 DNANet 76.6 23.8 0.610 0.799 0.770 0.713 ABCNet 67.6 19.7 0.599 0.699 0.727 0.682 Ours 96.8 36.8 0.634 0.755 0.789 0.750 
下载: 导出CSV
表 2 不同方法的参数量和检测时间比较
Table 2 Comparison of parameter count and detection time among different methods
AGPC DNA FC3 ACM ABC LSPM Ours Parameter count 12423175 4696517 7043282 530335 73508047 31583202 5205458 Detection time/s 51 25 14 10 151 77 21
下载: 导出CSV
表 3 不同模块组合检测结构的指标对比
Table 3 Comparison of metrics for different module combinations in detection structures
Biformer1 Biformer2 Biformer3 AFM IoU Fmeasure × × × × 0.264 0.341 × × × √ 0.549 0.660 √ × × √ 0.556 0.673 √ √ × √ 0.573 0.680 √ √ √ × 0.592 0.693 √ √ √ √ 0.634 0.750
下载: 导出CSV
-
[1] KOU R, WANG C, PENG Z, et al. Infrared small target segmentation networks: A survey[J]. Pattern Recognition, 2023, 143: 109788. DOI: 10.1016/j.patcog.2023.109788
[2] 崔晨辉, 蔺素珍, 李大威, 等. 基于孪生网络和Transformer的红外弱小目标跟踪方法[J]. 计算机应用, 2024, 44(2): 563-571. CUI C H, LIN S Z, LI D W, et al. Infrared weak target tracking method based on twin network and transformer[J]. Journal of Computer Applications, 2024, 44(2): 563-571.
[3] 蔺素珍, 张海松, 禄晓飞, 等. RBNSM: 一种复杂背景下红外弱小目标检测新方法[J]. 红外技术, 2022, 44(7): 667-675. http://hwjs.nvir.cn/article/id/fb6071f5-08ab-4944-b4a1-9921a68c5948 LIN S Z, ZHANG H S, LU X F, et al. RBNSM: A new method for infrared dim and small target detection in complex backgrounds[J]. Infrared Technology, 2022, 44(7): 667-675. http://hwjs.nvir.cn/article/id/fb6071f5-08ab-4944-b4a1-9921a68c5948
[4] LIANG X, LIU L, LUO M, et al. Robust infrared small target detection using hough line suppression and rank-hierarchy in complex backgrounds[J]. Infrared Physics & Technology, 2022, 120: 103893.
[5] LIU F, GAO C, CHEN F, et al. Infrared small and dim target detection with transformer under complex backgrounds[J]. IEEE Transactions on Image Processing, 2023, 32: 5921-5932. DOI: 10.1109/TIP.2023.3326396
[6] DENG L, ZHANG J, XU G, et al. Infrared small target detection via adaptive M-estimator ring top-hat transformation[J]. Pattern Recognition, 2021, 112: 107729.
[7] ZHANG T, PENG Z, WU H, et al. Infrared small target detection via self-regularized weighted sparse model[J]. Neurocomputing, 2021, 420: 124-148.
[8] LU Y, HUANG S, ZHAO W, et al. Sparse representation based infrared small target detection via an online-learned double sparse background dictionary[J]. Infrared Physics & Technology, 2019, 99: 14-27.
[9] CHEN C L P, LI H, WEI Y, et al. A local contrast method for small infrared target detection[J]. IEEE Transactions on Geoscience and Remote Sensing, 2013, 52(1): 574-581.
[10] Pustokhina I V, Pustokhin D A, Vaiyapuri T, et al. An automated deep learning based anomaly detection in pedestrian walkways for vulnerable road users safety[J]. Safety Science, 2021, 142: 105356.
[11] XIE S H, ZHANG W Z, CHENG P, et al. YOLOv4 fire and smoke detection model with embedded channel attention[J]. Chinese Journal of Liquid Crystal & Displays, 2021, 36(10): 1445-1453.
[12] DAI Y, WU Y, ZHOU F, et al. Asymmetric contextual modulation for infrared small target detection[C]//Proceedings of the IEEE/CVF Winter Conference on Applications of Computer Vision, 2021: 950-959.
[13] DAI Y, WU Y, ZHOU F, et al. Attentional local contrast networks for infrared small target detection[J]. IEEE Transactions on Geoscience and Remote Sensing, 2021, 59(11): 9813-9823.
[14] WANG H, ZHOU L, WANG L, et al. Miss detection vs. false alarm: Adversarial learning for small object segmentation in infrared images [C]// Proceedings of the IEEE/CVF International Conference on Computer Vision, 2019: 8509-8518.
[15] LI B, XIAO C, WANG L, et al. Dense nested attention network for infrared small target detection[J]. IEEE Transactions on Image Processing, 2022, 32: 1745-1758.
[16] CARION N, MASSA F, SYNNAEVE G, et al. End-to-end object detection with transformers[J/OL]. arXiv: 2005.12872, https://arxiv.org/abs/2005.12872.
[17] ZHANG F, LIN S, XIAO X, et al. Global attention network with multiscale feature fusion for infrared small target detection[J]. Optics & Laser Technology, 2024, 168: 110012.
[18] HUANG L, DAI S, HUANG T, et al. Infrared small target segmentation with multiscale feature representation[J]. Infrared Physics & Technology, 2021, 116: 103755.
[19] PAN P, WANG H, WANG C, et al. ABC: attention with bilinear correlation for infrared small target detection[J]. arXiv preprint arXiv: 2303.10321, 2023.
[20] ZHANG H, WU C, ZHANG Z, et al. Resnest: split-attention networks[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2022: 2736-2746.
[21] ZHU L, WANG X, KE Z, et al. BiFormer: vision transformer with bi-level routing attention[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 2023: 10323-10333.
[22] ZHANG M, YUE K, ZHANG J, et al. Exploring feature compensation and cross-level correlation for infrared small target detection[C]// Proceedings of the 30th ACM International Conference on Multimedia, 2022: 1857-1865.
[23] ZHANG T, LI L, CAO S, et al. AGPCNet: Attention-guided pyramid context networks for detecting infrared small target under complex background[J]. IEEE Transactions on Aerospace and Electronic Systems, 2023, 59(4): 4250-4261.
计量
- 文章访问数: 34
- HTML全文浏览量: 7
- PDF下载量: 11
