Article Details

Title · Authors · Abstract · Keywords · DOI · References

YOLO-IRLight: Lightweight UAV Infrared Small Target Detection Algorithm Based on YOLOv8

NI Mengqi, CHEN Kaiyuan

(North China University of Water Resources and Electric Power, Zhengzhou 450046, China)

Section：Artificial Intelligence · Vol. 2025, Issue 8 · Pages：46-53,60 · 2096-4706（2025）08-0046-09 · CLC number：TP3191.4；TP301.6 · Document code：A · DOI：10.19850/j.cnki.2096-4706.2025.08.010

Abstract

To address the issues of low detection accuracy and high computational load in infrared small target detection from UAV aerial perspectives, a lightweight infrared small target detection model, YOLO-IRLight, is proposed based on YOLOv8s. This model introduces the EMA (Efficient Multiscale Attention) Attention Mechanism to enhance feature extraction capabilities. A PConv-C2f module is added to the neck of the network to reduce computational load and fuse scale sequence features, and a P2 detection layer is incorporated to optimize the network structure, thereby improving small target detection performance. A novel lightweight detection head, Group-Detect, is designed, and the NWD (Normalized Gaussian Wasserstein Distance) loss function is incorporated into the loss function of the model in a linear combination manner. Experimental results on the open dataset show that compared to the original YOLOv8s, the proposed model improves detection accuracy (mAP@0.5) by 1.7%, reduces the number of parameters by 45.9%, decreases computational complexity (GFLOPs) by 33.5%, and increases F1 score by 0.9%. The improved algorithm significantly outperforms traditional algorithms, with notable improvements in detection accuracy compared to current mainstream algorithms.

Keywords

Small Target Detection; infrared target; lightweight; YOLOv8; network optimization

References

  1. [1] Li Wenbo, Wang Qi, noble A review of deep learning based infrared small target detection algorithms [J]. Laser and Infrared, 2023, 53 (10): 1476-1484
  2. [2] Song Wanni, Yang Benchen, Jin Haibo Infrared small target detection based on improved weighted local contrast [J]. Laser and Infrared, 2023, 53 (6): 963-969
  3. [3] YOUNSI M，DIAF M，SIARRY P. Automatic Multiple Moving Humans Detection and Tracking in Image Sequences Taken from a Stationary Thermal Infrared Camera [J/OL].Expert Systems with Applications，2020，146：113171[2024-10-05]. https://doi.org/10.1016/j.eswa.2019.113171.
  4. [4] QI M，LIU L，ZHUANG S，et al. FTC-Net: Fusion of Transformer and CNN Features for Infrared Small Target Detection [J].IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing，2022，15：8613-8623.
  5. [5] JIANG C C，REN H Z，YANG H，et al. M2FNet: Multi-Modal Fusion Network for Object Detection from Visible and Thermal Infrared Images [J/OL].International Journal of Applied Earth Observation and Geoinformation，2024，130：103918[2024-10-23].https://doi.org/10.1016/j.jag.2024.103918.
  6. [6] LIANG M，JI T. Research on Unmanned Infrared Night Vision System Based on Improved Faster R-CNN Target Detection Algorithm [C]//2022 IEEE 5th Advanced Information Management，Communicates，Electronic and Automation Control Conference (IMCEC).Chongqing：IEEE，2022：863-869.
  7. [7] WANG K，WANG Y，ZHANG S，et al. SLMS-SSD: Improving the Balance of Semantic and Spatial Information in Object Detection [J/OL].Expert Systems with Applications，2022，206：117682[2024-10-11]. https: //doi.org/10.1016/j.eswa.2022.117682.
  8. [8] LU J，HUANG T，ZHANG Q，et al. A Lightweight Vehicle Detection Network Fusing Feature Pyramid and Channel Attention [J/OL].Internet of Things，2024，26：101116[2024-10-18].https://doi.org/10.1016/j.iot.2024.101166.
  9. [9] WANG H，MO H，LU S，et al. Electrolytic Capacitor Surface Defect Detection Based on Deep Convolution Neural Network [J/OL].Journal of King Saud University-Computer and Information Sciences，2024，36（2）：101935[2024-10-23].https://doi.org/10.1016/j.jksuci.2024.101935.
  10. [10] ZHENG H，SUN Y，LIU X，et al. Infrared Image Detection of Substation Insulators Using an Improved Fusion Single Shot Multibox Detector [J].IEEE Transactions on Power Delivery，2020，36（6）：3351-3359.
  11. [11] XINGANG M，SHUAI L，XIAO Z. YOLO-FR: A YOLOv5 Infrared Small Target Detection Algorithm Based on Feature Reassembly Sampling Method [J].Sensors，2023，23（5）：2710-2710.
  12. [12] ZHU J X，QIN C，CHOI D M. YOLO-SDLUWD: YOLOv7-based Small Target Detection Network for Infrared Images in Complex Backgrounds [J/OL].Digital Communications and Networks，2023[2024-10-27].https://doi.org/10.1016/j.dcan.2023.11.001.
  13. [13] SUN S，MO B，XU J，et al. Multi-YOLOv8: An Infrared Moving Small Object Detection Model Based on YOLOv8 for Air Vehicle [J/OL].Neurocomputing，2024，588：127685[2024-11-05].https://doi.org/10.1016/j.neucom.2024.127685.
  14. [14] CHEN J，KAO S，HE H，et al. Run，Don't Walk: Chasing Higher FLOPS for Faster Neural Networks [C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.Vancouver：IEEE，2023：12021-12031.
  15. [15] KANG M，TING C M，TING F F，et al. ASF-YOLO: A Novel YOLO Model with Attentional Scale Sequence Fusion for Cell Instance Segmentation [J/OL].Image and Vision Computing，2024，147：105057[2024-11-06].https://doi.org/10.1016/j.imavis.2024.105057.
  16. [16] WANG J，XU C，YANG W，et al. A Normalized Gaussian Wasserstein Distance for Tiny Object Detection [J/OL].arXiv:2110.13389 [cs.CV].[2024-10-23].2021.https://arxiv.org/abs/2110.13389v1.
  17. [17] HE S，GIRSHICK R，DOLLÁR P，et al. Aggregated Residual Transformations for Deep Neural Networks [C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition.Honolulu：IEEE，2017：1492-1500.
  18. [18] TAN M X，PANG R M，LE Q V. EfficientDet: Scalable and Efficient Object Detection [C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.Seattle：IEEE，2020：10781-10790.
  19. [19] LIU W，ANGUELOV D，ERHAN D，et al. SSD: Single Shot Multibox Detector [C]//Computer Vision-ECCV 2016: 14th European Conference，Amsterdam，The Netherlands，October 11–14，2016，Proceedings，Part I.Amsterdam：Springer International Publishing，2016：21-37.
  20. [20] REN S，HE K，GIRSHICK R，et al. Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks [J].IEEE Transactions on Pattern Analysis and Machine Intelligence，2016，39（6）：1137-1149.
  21. [21] LIN T Y，GOYAL P，GIRSHICK R，et al. Focal Loss for Dense Object Detection [C]//Proceedings of the IEEE International Conference on Computer Vision. Venice：IEEE，2017：2980-2988.
  22. [22] ZHOU X，WANG D，KRÄHENBÜHL P. Objects as Points [J/OL].arXiv:1904.07850 [cs.CV].（2019-04-16）[2024-10-05].https://arxiv.org/abs/1904.07850v1.
  23. [23] Ultralytics. Comprehensive Guide to Ultralytics YOLOv5 [EB/OL].（2022-11-22）[2024-10-07].https://docs.ultralytics.com/yolov5/.
  24. [24] WANG C Y，BOCHKOVSKIY A，LIAO H Y M. YOLOv7: Trainable Bag-of-Freebies Sets New State-of-the-Art for Real-Time Object Detectors [C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition.Vancouver：IEEE，2023：7464-7475.
  25. [25] Ultralytics. Ultralytics YOLOv8 [EB/OL].（2023-01-10）[2024-10-09].https://github.com/ultralytics/ultralytics.
  26. [26] Ultralytics. Ultralytics Solutions: Harness YOLO11 to Solve Real-World Problems [EB/OL].（2024-09-30）[2024-10-15].https://docs.ultralytics.com/zh.

About the authors

Ni Mengqi (1999-), female, Han ethnicity, from Luoyang, Henan Province, is currently pursuing a master's degree in computer target recognition and tracking; Chen Kaiyuan (2000-), female, Han ethnicity, from Zhengzhou, Henan Province, is currently pursuing a master's degree in defect detection and natural language processing.