# Research on Algorithm for Improving Infrared Image Defect Segmentation of Power Equipment

^{*}

## Abstract

**:**

## 1. Introduction

## 2. Countermeasure Datasets for Learning Attack Strategies

## 3. CS_DeeplabV3+ Model

#### 3.1. Deeplab V3+ Model

#### 3.2. CS_DeeplabV3+Model

#### 3.3. Improving Attention Mechanism Module

_{0}∈RC/r × C, W

_{1}∈RC × C/r, W

_{0}, and W

_{1}denote the two-layer network parameters of the MLP, and the final output of the channel attention is F’.

#### 3.4. Semantic Segmentation Feature Enhancement Module

#### 3.5. Improved Loss Function

_{t}represents the probability predicted by the model, and weight γ represents the rate of decline in the weight of the adjusted sample. When γ is set to 0, the focus loss function degrades to a cross-entropy loss function. As γ increases, so does the adjustment factor. The original focus loss function achieves the target training by suppressing the loss value of the sample to varying degrees. Suppression causes the weight of the sample loss value to decrease between 0 and 1. The weight of the simple sample is close to 0, and the weight of the difficult sample is close to 1, but during the semantic segmentation process, we need to correctly classify each pixel. Therefore, while retaining the loss value weight of easy samples, increasing the loss value weight of difficult samples is more suitable for semantic segmentation.

## 4. Experimental Evaluation and Discussion

#### 4.1. Dataset

#### 4.2. Experimental Evaluation Indicators

#### 4.3. Comparison Experiment and Analysis

## 5. Conclusions

## Author Contributions

## Funding

## Data Availability Statement

## Acknowledgments

## Conflicts of Interest

## References

- Cao, H.; Zhang, Y.; Zhu, Q.; Zhao, Q. On-line detection and diagnosis of transformer thermal fault based on infrared imaging technology. Lab. Res. Explor.
**2012**, 31, 30–32+40. [Google Scholar] - Li, T. Application of accurate infrared temperature measurement in condition maintenance of power equipment. High Volt. Appar.
**2022**, 56, 246–251. [Google Scholar] [CrossRef] - Liu, X.; Ma, J. The Application of the ubiquitous power Internet of Things in the state monitoring of power equipment. Power Syst. Prot. Control
**2020**, 48, 69–75. [Google Scholar] [CrossRef] - Zhang, Z.; Li, K.; Liu, J.; Jia, B.; Wang, P.; Geng, J. Live detection technology of deteriorated insulator based on UAV inspection platform. Sci. Technol. Eng.
**2020**, 20, 8616–8621. [Google Scholar] [CrossRef] - Jin, L.; Tian, Z.; Gao, K.; Xiao, R. Insulator contamination level recognition based on infrared and visible light image information fusion. Proc. CSEE
**2016**, 36, 3682–3691. [Google Scholar] [CrossRef] - Peng, Z.; Chen, Y.; Pu, T.; Wang, Y.; He, Y. Review of image denoising methods based on sparse representation and regular constraints. Data Acquis. Process.
**2018**, 33, 1–11. [Google Scholar] [CrossRef] - Zhang, N.; Zhang, Y.; Ding, W. A Review of Classical Image Denoising Methods. Chem. Autom. Instrum.
**2021**, 48, 409–412, 423. [Google Scholar] [CrossRef] - Tang, C.; Tao, Q.; Huang, M. Discussion on the application of infrared temperature measurement technology in substation operation and maintenance. Shi He Zi Sci. Technol.
**2020**, 06, 14–15. [Google Scholar] [CrossRef] - Chen, D.; Tang, W.; Niu, Z. Fault diagnosis method of infrared image of power equipment based on deep learning. Guangdong Electr. Power
**2021**, 34, 97–105. [Google Scholar] [CrossRef] - Liu, H.; Yu, C.; Wu, H.; Duan, Z.; Yan, G. A new hybrid ensemble deep reinforcement learning model for wind speed short term forecasting. Energy
**2020**, 202, 117794. [Google Scholar] [CrossRef] - Shao, Z.; Zhang, Z.; Wei, W.; Wang, F.; Xu, Y.; Cao, X.; Jensen, C.S. Decoupled dynamic spatial-temporal graph neural network for traffic forecasting. arXiv
**2022**, arXiv:2206.09112. [Google Scholar] [CrossRef] - Qi, C.; Li, Q.; Liu, Y.; Ni, J.; Ma, R.; Xu, Z. Infrared image segmentation based on multi-information fused fuzzy clustering method for electrical equipment. Int. J. Adv. Robot. Syst.
**2020**, 17, 1–18. [Google Scholar] [CrossRef] - Shu, J.; He, J.; Li, L. MSIS: Multispectral Instance Segmentation Method for Power Equipment. Comput. Intell. Neurosci.
**2022**, 2022, 2864717. [Google Scholar] [CrossRef] [PubMed] - Long, J.; Shelhamer, E.; Darrell, T. Fully convolutional networks for semantic segmentation. In Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Boston, MA, USA, 7–12 June 2015; pp. 3431–3440. [Google Scholar]
- Badrinarayanan, V.; Kendall, A.; Cipolla, R. SegNet: A Deep Convolutional Encoder-Decoder Architecture for Image Segmentation. IEEE Trans Pattern Anal. Mach. Intell.
**2017**, 39, 2481–2495. [Google Scholar] [CrossRef] [PubMed] - Ronneberger, O.; Fischer, P.; Brox, T. U-Net: Convolutional Networks for Biomedical Image Segmentation. In Proceedings of the Medical Image Computing and Computer-Assisted Intervention—MICCAI, Munich, Germany, 5–9 October 2015; Volume 9351, pp. 234–241. [Google Scholar]
- Usamentiaga, R.; Lema, D.G.; Pedrayes, O.D.; Garcia, D.F. Automated surface defect detection in metals: A comparative review of object detection and semantic segmentation using deep learning. IEEE Trans. Ind. Appl.
**2022**, 58, 4203–4213. [Google Scholar] [CrossRef] - Chen, L.-C.; Zhu, Y.; Papandreou, G.; Schroff, F.; Adam, H. Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation. In Proceedings of the Computer Vision—ECCV, Munich, Germany, 8–14 September 2018; Volume 11211, pp. 833–851. [Google Scholar]
- Chen, L.-C.; Papandreou, G.; Kokkinos, I.; Murphy, K.; Yuille, A.L. DeepLab: Semantic Image Segmentation with Deep Convolutional Nets, Atrous Convolution, and Fully Connected CRFs. IEEE Trans. Pattern Anal. Mach. Intell.
**2018**, 40, 834–848. [Google Scholar] [CrossRef] [Green Version] - Liu, S.; Yin, F. Image segmentation method based on graph secsion and its new progress. Acta Autom. Sin.
**2012**, 38, 911–922. [Google Scholar] [CrossRef] - Zeng, J.; Huang, S. Comparison and analysis of typical image edge detection operators. J. Hebei Norm. Univ. (Nat. Sci. Ed.)
**2020**, 44, 295–301. [Google Scholar] [CrossRef] - Dang, W. Research and application of image region segmentation algorithm. Anhui Univ. Sci. Technol.
**2018**, 12, 76. [Google Scholar] [CrossRef] - Jia, X.; Zhang, Y.; Wu, B.; Ma, K.; Wang, J.; Cao, X. LAS-AT: Adversarial training with learnable attack strategy. In Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition, New Orleans, LA, USA, 18–24 June 2022; pp. 13398–13408. [Google Scholar]
- Meng, L.; Xu, L.; Guo, J. A Semantic Segmentation Algorithm for MobileNetV2 Network Based on Improvement. Chin. J. Electron.
**2020**, 48, 1769–1776. [Google Scholar] [CrossRef] - Chollet, F. Xception: Deep Learning with Depthwise Separable Convolutions. In Proceedings of the 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR), Honolulu, HI, USA, 21–26 July 2017; pp. 1800–1807. [Google Scholar]
- Guo, L.; Zhang, T.; Sun, W.; Guo, J. Pictorial Semantic Description Algorithm for Fusion Spatial Attention Mechanism. Adv. Laser Optoelectron.
**2021**, 58, 313–322. [Google Scholar] [CrossRef] - Hou, T.; Zhao, J.; Qiang, Y.; Sanhu, W.; Pan, W. The CRF 3D-UNet lung nodule segmentation network. Comput. Eng. Des.
**2020**, 41, 1663–1669. [Google Scholar] [CrossRef] - Li, K.; Han, B.; Zhang, J. Detection and Recognition of Traffic Signs Based on Conditional Random Airport and Multi-scale Convolutional Neural Network. Comput. Appl.
**2020**, 41, 1663–1669. [Google Scholar]

**Figure 8.**Original drawing and segmentation effects for 3 different models: (

**a**) original drawing; (

**b**) DeeplabV3+; (

**c**) CBAM_DeeplabV3+; and (

**d**) CS_DeeplabV3+.

**Figure 9.**Network prediction effect diagram: (

**a**) original drawing; (

**b**) U-Net; (

**c**) Deeplabv3+; and (

**d**) CS_Deeplabv3+.

Module | MPA | MIOU |
---|---|---|

DeeplabV3+ | 0.794 | 0.699 |

CBAM_DeeplabV3+ | 0.824 | 0.738 |

CS_DeeplabV3+ | 0.850 | 0.772 |

Module | MPA | MIOU |
---|---|---|

U-Net | 0.714 | 0.615 |

DeeplabV3+ | 0.794 | 0.699 |

CS_DeeplabV3+ | 0.850 | 0.772 |

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content. |

© 2023 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

## Share and Cite

**MDPI and ACS Style**

Zhang, J.; Zhu, W.
Research on Algorithm for Improving Infrared Image Defect Segmentation of Power Equipment. *Electronics* **2023**, *12*, 1588.
https://doi.org/10.3390/electronics12071588

**AMA Style**

Zhang J, Zhu W.
Research on Algorithm for Improving Infrared Image Defect Segmentation of Power Equipment. *Electronics*. 2023; 12(7):1588.
https://doi.org/10.3390/electronics12071588

**Chicago/Turabian Style**

Zhang, Jingwen, and Wu Zhu.
2023. "Research on Algorithm for Improving Infrared Image Defect Segmentation of Power Equipment" *Electronics* 12, no. 7: 1588.
https://doi.org/10.3390/electronics12071588