Skip to main content

Anchor-Free RetinaNet: Anchor-Free Based Insulator Status Detection Method

  • Conference paper
  • First Online:
The Proceedings of the 17th Annual Conference of China Electrotechnical Society (ACCES 2022)

Part of the book series: Lecture Notes in Electrical Engineering ((LNEE,volume 1012))

Included in the following conference series:

  • 879 Accesses

Abstract

When UAVs are used to detect the status of insulators, the appearance of insulators may change greatly because of the great variation of shooting angle and distance. Thus, due to the inefficient hand-craft anchor, the existing anchor-based object detection methods are difficult to perform robust detection of the status of insulators. Anchor-free object detection methods, in which samples needed for training are automatically allocated to each feature pyramid layer, can overcomes the disadvantage of poor detection with large morphological changes by hand-craft anchor. In this paper, we propose anchor-free RetinaNet to improve the accuracy of insulators status detection with different morphological changes. Specifically, anchor-free RetinaNet uses sample generation of spatial dimension to automatically generate positive samples and uses unified focal loss to introduce regression task into classification to enhance the model’s detection accuracy. Experiments on the public insulator status detection dataset show that the proposed method can improve the training efficiency and performance obviously.

This is a preview of subscription content, log in via an institution to check access.

Access this chapter

Chapter
USD 29.95
Price excludes VAT (USA)
  • Available as PDF
  • Read on any device
  • Instant download
  • Own it forever
eBook
USD 349.00
Price excludes VAT (USA)
  • Available as EPUB and PDF
  • Read on any device
  • Instant download
  • Own it forever
Softcover Book
USD 449.99
Price excludes VAT (USA)
  • Compact, lightweight edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide - see info
Hardcover Book
USD 449.99
Price excludes VAT (USA)
  • Durable hardcover edition
  • Dispatched in 3 to 5 business days
  • Free shipping worldwide - see info

Tax calculation will be finalised at checkout

Purchases are for personal use only

Institutional subscriptions

Similar content being viewed by others

References

  1. Huang, D., Wang, Y., Hu, A., et al.: An edge recognition method for insulator state based on multi-dimension feature fusion. Electric Power 55(1), 133–141 (2022). (in Chinese)

    Google Scholar 

  2. Guo, T., Yang, H., Shi, L., et al.: Self-explosion defect identification of insulator based on faster RCNN. Insulators and Surge Arresters 3, 183–189 (2019). (in Chinese)

    Google Scholar 

  3. Ren, S., He, K., Girshick, R., et al.: Faster R-CNN: towards real-time object detection with region proposal networks. IEEE Trans. Pattern Anal. Mach. Intell. 39(6), 1137–1149 (2017)

    Article  Google Scholar 

  4. Hou, C., Zhang, H., Zhang, W., et al.: Recognition method for faults of insulators on transmission lines. Proceedings of the CSU-EPSA 6, 1–6 (2019). (in Chinese)

    Google Scholar 

  5. Gao, Y., Tian, L., Du, Q., et al.: overheating defect detection of composite insulator based on mask R-CNN. Electric Power 54(1), 135–141 (2021). (in Chinese)

    Google Scholar 

  6. He, K., Gkioxari, G., Dollár, P., et al.: Mask r-cnn. In: Proceedings of the IEEE International Conference on Computer Vision, pp. 2961–2969 (2017)

    Google Scholar 

  7. Liu, Y., Wang, S., Qing, Y., et al.: insulator defect detection based on EfficientDet and binocular camera. Electric Power 54(2), 156–163 (2021). (in Chinese)

    Google Scholar 

  8. Tan, M., Pang, R., Le, Q.V.: Efficientdet: scalable and efficient object detection. In: Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, 10781–10790 (2020)

    Google Scholar 

  9. Lin, T.Y., Goyal, P., Girshick, R., et al.: Focal loss for dense object detection. In: Proceedings of the IEEE International Conference on Computer Vision, pp. 2980–2988 (2017)

    Google Scholar 

  10. Lin, T.Y., Dollár, P., Girshick, R., et al.: Feature pyramid networks for object detection. In: Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition, pp. 2117–2125. IEEE, Las Vegas. America (2017)

    Google Scholar 

  11. Dou, Z., Gao, K., Zhang, X.: Improving performance and adaptivity of anchor-based detector using differentiable an-choring with efficient target generation. IEEE Trans. Image Process. 30, 712–724 (2020)

    Article  Google Scholar 

  12. Tao, X., Zhang, D., Wang, Z., Liu, X., Zhang, H., De, X.: Detection of power line insulator defects using aerial images analyzed with convolutional neural networks. IEEE Trans. Syst. Man Cybern.: Syst. 50(4), 1486–1498 (2020). https://doi.org/10.1109/TSMC.2018.2871750

    Article  Google Scholar 

  13. Lin, T.-Y., et al.: Microsoft coco: common objects in context. In: Fleet, D., Pajdla, T., Schiele, B., Tuytelaars, T. (eds.) Computer Vision – ECCV 2014: 13th European Conference, Zurich, Switzerland, September 6-12, 2014, Proceedings, Part V, pp. 740–755. Springer International Publishing, Cham (2014). https://doi.org/10.1007/978-3-319-10602-1_48

    Chapter  Google Scholar 

Download references

Acknowledgments

This work is supported by the Science and Technology Project of State Grid Ningxia Electric Power Co. Ltd (Project No. 5229WZ2000B3).

Author information

Authors and Affiliations

Authors

Corresponding author

Correspondence to Wendong Guo .

Editor information

Editors and Affiliations

Rights and permissions

Reprints and permissions

Copyright information

© 2023 Beijing Paike Culture Commu. Co., Ltd.

About this paper

Check for updates. Verify currency and authenticity via CrossMark

Cite this paper

Guo, W. et al. (2023). Anchor-Free RetinaNet: Anchor-Free Based Insulator Status Detection Method. In: Yang, Q., Li, J., Xie, K., Hu, J. (eds) The Proceedings of the 17th Annual Conference of China Electrotechnical Society. ACCES 2022. Lecture Notes in Electrical Engineering, vol 1012. Springer, Singapore. https://doi.org/10.1007/978-981-99-0357-3_14

Download citation

  • DOI: https://doi.org/10.1007/978-981-99-0357-3_14

  • Published:

  • Publisher Name: Springer, Singapore

  • Print ISBN: 978-981-99-0356-6

  • Online ISBN: 978-981-99-0357-3

  • eBook Packages: EngineeringEngineering (R0)

Publish with us

Policies and ethics