Skip to content
2000
Volume 18, Issue 7
  • ISSN: 2666-2558
  • E-ISSN: 2666-2566

Abstract

Introduction

To address the challenges of low visibility, object recognition difficulties, and low detection accuracy in foggy weather, this paper introduces the WVIT-YOLO real-time fog detection model, built on the YOLOv5 framework. The NVIT-Net backbone network, incorporating NTB and NCB modules, enhances the model's ability to extract both global and local features from images.

Methods

An efficient convolutional C3_DSConv module is designed and integrated with channel attention mechanisms and ShuffleAttention at each upsampling stage, improving the model's computational speed and its ability to detect small and blurry objects. The Wise-IOU loss function is utilized during the prediction stage to enhance the model's convergence efficiency.

Results and Discussion

Experimental results on the publicly available RTTS dataset for vehicle detection in foggy conditions demonstrate that the WVIT-YOLO model achieves a 3.2% increase in precision, a 9.5% rise in recall, and an 8.6% improvement in mAP50 compared to the baseline model. Furthermore, WVIT-YOLO shows a 9.5% and 8.6% mAP50 improvement over YOLOv7 and YOLOv8, respectively. For detecting small and blurry objects in foggy conditions, the model demonstrates approximately a 5% improvement over the benchmark, significantly enhancing the detection network's generalization ability under foggy conditions.

Conclusion

This advancement is crucial for improving vehicle safety in such weather. The code is available at https://github.com/QinghuaZhang1/mode.

© 2025 Bentham Science Publishers
Loading

Article metrics loading...

/content/journals/rascs/10.2174/0126662558331725241024101310
2024-11-04
2025-11-02

  • From This Site

    /content/journals/rascs/10.2174/0126662558331725241024101310
    dcterms_title,dcterms_subject,pub_keyword
    -contentType:Contributor -contentType:Concept -contentType:Institution
    10
    5
Loading full text...

Full text loading...

References

  1. GuoC. Zero-reference deep curve estimation for low-light image enhancement.2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)13-19 Jun, 2020, Seattle, WA, USA, 2020, pp. 1777-1786.10.1109/CVPR42600.2020.00185
     [Google Scholar]
  2. CaiY. DaiL. WangH. ChenL. LiY. DLnet with training task conversion stream for precise semantic segmentation in actual traffic scene.IEEE Transac. Neural Net. Learning Sys.2021336443645710.1109/TNNLS.2021.3080261
     [Google Scholar]
  3. LiuX. LinY. YOLO-GW: quickly and accurately detecting pedestrians in a foggy traffic environment.Sensors (Basel)20232312553910.3390/s2312553937420706
     [Google Scholar]
  4. LiuX. MaY. ShiZ. ChenJ. GridDehazeNet: Attention-Based Multi-Scale Network for Image Dehazing.2019 IEEE/CVF International Conference on Computer Vision (ICCV)27 Oct - 02 Nov, 2019, Seoul, Korea (South), 2019, pp. 7313-7322.10.1109/ICCV.2019.00741
     [Google Scholar]
  5. DongH. Multi-scale boosted dehazing network with dense feature fusion.2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)13-19 Jun, 2020, Seattle, WA, USA, 2020, pp. 2154-2164.10.1109/CVPR42600.2020.00223
     [Google Scholar]
  6. QinX. WangZ. BaiY. XieX. JiaH. “FFA-Net: Feature fusion attention network for single image dehazing,” in Proc. AAAI Conf.Artif. Intell.20203471190811915
     [Google Scholar]
  7. LiB. PengX. WangZ. XuJ. FengD. AOD-Net: All-in-one dehazing network.2017 IEEE International Conference on Computer Vision (ICCV)22-29 Oct, 2017, Venice, Italy, 2017, pp. 4780-4788.10.1109/ICCV.2017.511
     [Google Scholar]
  8. SindagiV.A. OzaP. Prior-based domain adaptive object detection for hazy and rainy conditions.Computer Vision – ECCV 2020Glasgow, U.K.Springer2020763780
     [Google Scholar]
  9. HuangS-C. LeT-H. JawD-W. DSNet: Joint semantic learning for object detection in inclement weather conditions.IEEE Trans. Pattern Anal. Mach. Intell.20214382623263332149681
     [Google Scholar]
  10. ChenY. LiW. SakaridisC. DaiD. an GoolL.V. Domain adaptive faster R-CNN for object detection in the wild.2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition18-23 Jun, 2018, Salt Lake City, UT, USA, 2018, pp. 3339-3348.10.1109/CVPR.2018.00352
     [Google Scholar]
  11. ZhangS. TuoH. HuJ. JingZ. Domain adaptive YOLO for one-stage cross-domain detectionProc. Asian Conf. Mach. Learn.2021157785797
     [Google Scholar]
  12. KansalI. KasanaS.S. Improved color attenuation prior based image de-fogging technique.Multimedia Tools Appl.20207917-18120691209110.1007/s11042‑019‑08240‑6
     [Google Scholar]
  13. CaiB. XuX. JiaK. QingC. TaoD. Dehazenet: An end-to-end system for single image haze removal.IEEE Trans. Image Process.201625115187519810.1109/TIP.2016.259868128873058
     [Google Scholar]
  14. LinH-Y. LinC-J. Using a hybrid of fuzzy theory and neural network filter for single image dehazing.Appl. Intell.20174741099111410.1007/s10489‑017‑0942‑z
     [Google Scholar]
  15. EigenD. Restoring an image taken through a window covered with dirt or rain.ICCV '13: Proceedings of the 2013 IEEE International Conference on Computer VisionIEEE Computer Society2013
     [Google Scholar]
  16. LiuL. OuyangW. WangX. FieguthP. ChenJ. LiuX. PietikäinenM. Deep learning for generic object detection: A survey.Int. J. Comput. Vis.2020128226131810.1007/s11263‑019‑01247‑4
     [Google Scholar]
  17. ZhengxiaZ.O.U. ZhenweiS.H.I. YuhongG.U.O. Object detection in 20 years: A surveyProceed. IEEE20191113257276
     [Google Scholar]
  18. DalalN. Histograms of oriented gradients for human detection.2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition25 May 2021, San Diego, USA, 2005, 886–893.10.1109/CVPR.2005.177
     [Google Scholar]
  19. KrizhevskyA. ImageNet classification with deep convolutional neural networks.Commun. ACM20126068490
     [Google Scholar]
  20. LeCunY. BengioY. HintonG. Deep learning.Nature2015521755343644410.1038/nature1453926017442
     [Google Scholar]
  21. GirshickR. DonahueJ. DarrellT. Rich feature hierarchies for accurate object detection and semantic segmentation.2014 IEEE Conference on Computer Vision and Pattern Recognition23-28 Jun, 2014, Columbus, OH, USA, 2014, pp. 580-587.
     [Google Scholar]
  22. GirshickR. Fast R-CNN.2015 IEEE International Conference on Computer Vision (ICCV)07-13 Dec, 2015, Santiago, Chile, 2015, pp. 1440-1448.
     [Google Scholar]
  23. RenS. HeK. GirshickR. Faster r-cnn: Towards real-time object detection with region proposal networks.Adv. Neural Inf. Process. Syst.2015201528
     [Google Scholar]
  24. LiuW. AnguelovD. ErhanD. Ssd: Single shot multibox detector, Computer Vision–ECCV 2016.14th European ConferenceOctober 11–14, 2016, Amsterdam, The Netherlands, pp. 21-37.
     [Google Scholar]
  25. RedmonJ. DivvalaS. GirshickR. You only look once: Unified, real-time object detection.2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)27-30 Jun, 2016, Las Vegas, NV, USA, 2016, pp. 779-788.
     [Google Scholar]
  26. BochkovskiyA. WangC.Y. LiaoH.Y.M. Yolov4: Optimal speed and accuracy of object detectionarXiv.2004.109342020
     [Google Scholar]
  27. FarhadiA. YOLOv3: An incremental improvementarXiv.1804.027672018
     [Google Scholar]
  28. WangC.Y. BochkovskiyA. LiaoH.Y.M. YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors.2023 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)17-24 Jun, 2023, Vancouver, BC, Canada, 2023, pp. 7464-7475.10.1109/CVPR52729.2023.00721
     [Google Scholar]
  29. LiY. YouS. BrownM.S. TanR.T. Haze visibility enhancement: A Survey and quantitative benchmarking.Comput. Vis. Image Underst.201716511610.1016/j.cviu.2017.09.003
     [Google Scholar]
  30. XU Y Review of video and image defogging algorithms and related studies on image restoration and enhancement.IEEE Access20154165188
     [Google Scholar]
  31. SinghD. KumarV. KUMAR V J.A comprehensive review of computational dehazing techniques.Arch. Comput. Methods Eng.20192651395141310.1007/s11831‑018‑9294‑z
     [Google Scholar]
  32. GuiJ. A comprehensive survey on image dehazing based on deep learning.ArXiv abs/2106.033232021
     [Google Scholar]
  33. Kaiming He Jian Sun Xiaoou Tang Single image haze removal using dark channel prior.IEEE Trans. Pattern Anal. Mach. Intell.201133122341235310.1109/TPAMI.2010.16820820075
     [Google Scholar]
  34. HeK. SunJ. TangX. Guided image filtering.IEEE Trans. Pattern Anal. Mach. Intell.20133561397140910.1109/TPAMI.2012.21323599054
     [Google Scholar]
  35. YangD. Proximal dehaze-net: A prior learning-based deep network for single image dehazing.Lecture Notes in Computer ScienceSpringer2018
     [Google Scholar]
  36. LiB. An all-in-one network for dehazing and beyond.arXiv:1707.065432017
     [Google Scholar]
  37. HasanM.K. DahalL. SamarakoonP.N. TusharF.I. MartíR. DSNet: Automatic dermoscopic skin lesion segmentation.Comput. Biol. Med.202012010373810.1016/j.compbiomed.2020.10373832421644
     [Google Scholar]
  38. ZHANG H IEEE Trans. Neural Netw. Learn. Syst.20193161794180731329133
     [Google Scholar]
  39. FangW. ZhangG. ZhengY. ChenY. Multi-task learning for UAV aerial object detection in foggy weather condition.Remote Sens. (Basel)20231518461710.3390/rs15184617
     [Google Scholar]
  40. GopalanR. Domain adaptation for object recognition: An unsupervised approach.ICCV '11: Proceedings of the 2011 International Conference on Computer VisionIEEE Computer Society2011
     [Google Scholar]
  41. MansourY. Domain adaptation:Learning bounds and algorithms.ArXiv abs/0902.34302009
     [Google Scholar]
  42. BEN- DAVID S A theory of learning from different domains.Mach. Learn.2010791151175
     [Google Scholar]
  43. ChenY. WangH. LiW. SakaridisC. DaiD. Van GoolL. Scale- aware domain adaptive Faster R-CNN.Int. J. Comput. Vis.202112972223224310.1007/s11263‑021‑01447‑x
     [Google Scholar]
  44. HeZ. Multi-adversarial faster-rcnn for unrestricted object detection. Proceedings of the IEEE/CVF International Conference on Computer Vision27 Oct - 02 Nov 2019, Seoul, Korea (South), pp. 6667-6676.
     [Google Scholar]
  45. ZhuX. Adapting object detectors via selective cross-domain alignment.Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)June 10–17, 2019, Nashville, Tennessee, pp.687-696.
     [Google Scholar]
  46. SaitoK. Strong-weak distribution alignment for adaptive object detection.2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)15-20 Jun, 2019, Long Beach, CA, USA, 2019, pp. 6949-6958.
     [Google Scholar]
  47. HnewaM. Multiscale domain adaptive yolo for cross-domain object detection.2021 IEEE International Conference on Image Processing (ICIP)19-22 Sept, 2021, Anchorage, AK, USA, 2021, pp. 3323-3327.
     [Google Scholar]
  48. WangY. Domain-specific suppression for adaptive object detection.2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR)20-25 Jun, 2021, Nashville, TN, USA, 2021, pp. 9598-9607.
     [Google Scholar]
  49. LiuW. RenG. YuR. GuoS. ZhuJ. ZhangL. Image-adaptive YOLO for object detection in adverse weather conditions.Proc. Conf. AAAI Artif. Intell.20223621792180010.1609/aaai.v36i2.20072
     [Google Scholar]
  50. LinT.Y. GoyalP. GirshickR. Focal loss for dense object detection.2017 IEEE International Conference on Computer Vision (ICCV)22-29 Oct, 2017, Venice, Italy, 2017, pp. 2999-3007.
     [Google Scholar]
  51. ShuaiK. JianwuZ. ZunjieZ. An improved YOLOv4 algorithm for pedestrian detection in complex visual scenes.Telecommun. Sci.2021378
     [Google Scholar]
  52. WangH. XuY. HeY. YOLOv5-Fog: A multiobjective visual detection algorithm for fog driving scenes based on improved YOLOv5.IEEE Trans. Instrum. Meas.20227111210.1109/TIM.2022.3216413
     [Google Scholar]
  53. TongZ. ChenY. XuZ. Wise-IoU: Bounding Box Regression Loss with Dynamic Focusing Mechanism.arXiv.2301.100512023
     [Google Scholar]
  54. LiJ. XiaX. LiW. Next-vit: Next generation vision transformer for efficient deployment in realistic industrial scenarios.ArXiv abs/2207.055012022
     [Google Scholar]
  55. LiuT. LiJ. CaiM. An Improved YOLOv3-SPP Algorithm for Image-Based Pothole Detection.Advances in Neural Networks – ISNN 2024: 18th International Symposium on Neural NetworksJuly 11–14, 2024, Weihai, China, pp. 328-335.
     [Google Scholar]
  56. SaoudL.S. NiuZ. SultanA. ADOD: Adaptive Domain-Aware Object Detection with Residual Attention for Underwater Environments.2023 21st International Conference on Advanced Robotics (ICAR)05-08 Dec, 2023, Abu Dhabi, United Arab Emirates, 2023, pp. 633-638.
     [Google Scholar]
  57. LiuJ. QiaoH. YangL. GuoJ. Improved lightweight YOLOv4 foreign object detection method for conveyor belts combined with CBAM.Appl. Sci. (Basel)20231314846510.3390/app13148465
     [Google Scholar]
  58. FuH. SongG. WangY. Improved YOLOv4 marine target detection combined with CBAM.Symmetry (Basel)202113462310.3390/sym13040623
     [Google Scholar]
  59. LinC.T. HuangS.W. WuY.Y. LaiS-H. GAN-based day-to-night image style transfer for nighttime vehicle detection.IEEE Trans. Intell. Transp. Syst.202122295196310.1109/TITS.2019.2961679
     [Google Scholar]
  60. YangL. ZhongJ. ZhangY. BaiS. LiG. YangY. ZhangJ. An improving faster-RCNN with multi-attention ResNet for small target detection in intelligent autonomous transport with 6G.IEEE Trans. Intell. Transp. Syst.20232477717772510.1109/TITS.2022.3193909
     [Google Scholar]
  61. YuG. ChangQ. LvW. PP-PicoDet: A better real-time object detector on mobile devices.arXiv.2111.009022021
     [Google Scholar]
  62. YuJ. JiangY. WangZ. UnitBox: An advanced object detection network.MM '16: Proceedings of the 24th ACM international conference on MultimediaOct 15-19, 2016, Melbourne, Australia, pp.516-520.
     [Google Scholar]
  63. ZhangY.F. RenW. ZhangZ. JiaZ. WangL. TanT. Focal and efficient IOU loss for accurate bounding box regression.Neurocomputing202250614615710.1016/j.neucom.2022.07.042
     [Google Scholar]
  64. ZhaoW. KangY. ChenH. ZhaoZ. ZhaiY. YangP. A target detection algorithm for remote sensing images based on a combination of feature fusion and improved anchor.IEEE Trans. Instrum. Meas.2022711810.1109/TIM.2022.3181927
     [Google Scholar]
  65. ZhengZ. WangP. LiuW. LiJ. YeR. RenD. Distance-IoU loss: Faster and better learning for bounding box regression.Proc. Conf. AAAI Artif. Intell.20203471299313000[C].10.1609/aaai.v34i07.6999
     [Google Scholar]
/content/journals/rascs/10.2174/0126662558331725241024101310
Loading
Real Time Object Detection Algorithm in Foggy Weather based on WVIT-YOLO Model
Recent Advances in Computer Science and Communications 18, E26662558331725 (2025); https://doi.org/10.2174/0126662558331725241024101310
/content/journals/rascs/10.2174/0126662558331725241024101310
/content/journals/rascs/10.2174/0126662558331725241024101310
Loading

Data & Media loading...


  • Article Type:     Research Article
Keyword(s): C3_DSConv; NVIT-Net; Real-time vehicle detection; vehicle safety; WIOU; WVIT-YOLO

Most Cited Most Cited RSS feed

This is a required field
Please enter a valid email address
Approval was a Success
Invalid data
An Error Occurred
Approval was partially successful, following selected items could not be processed due to error
Please enter a valid_number test