Detection of components on power lines using images captured by UAVs

• Pedro Arfux Pereira Cavalcante de Castro UFMS
• Wesley Nunes Gonçalves UFMS
• José Marcato Junior UFMS

DOI: https://doi.org/10.5753/wvc.2023.27536

Resumo

Preventive maintenance of power poles, including component inspection, is essential to ensure the continuous supply of electricity to cities. Despite its importance, the current practice of visually identifying issues demands considerable time and effort. To address this challenge, this study proposes the use of automatic methods for power pole inspection, aiming to reduce inspection time and predict potential component replacements. The investigation leverages convolutional neural networks based on Faster R-CNN to automatically evaluate power pole conditions. A dataset was generated from 848 images obtained from a repository available on the internet. The resulting neural network from the training process has an average precision of 98.7% with an Intersection over Union (IoU) of 50% and 98.2% with an IoU of 75% for high-voltage electrical network components.

Palavras-chave: Convolutional neural networks, Power line inspection, Unmanned aerial vehicle assisted inspection

Referências

R. Ferguson, W. Wilkinson, and R. Hill, “Electricity use and economic development,” Energy Policy, vol. 28, no. 13, pp. 923–934, 2000. [Online]. Available: [link].

V. N. Nguyen, R. Jenssen, and D. Roverso, “Automatic autonomous vision-based power line inspection: A review of current status and the potential role of deep learning,” International Journal of Electrical Power Energy Systems, vol. 99, pp. 107–120, 2018. [Online]. Available: [link].

T. V. Ferreira, “Estudo do trilhamento de núcleo em isoladores poliméricos,” Master’s thesis, Universidade Federal de Campina Grande, 2007.

L. Leite, J. Yanaguizawa, A. Shinohara, and G. Xavier, “Efeito da porosidade do cimento na tensão disruptiva de isoladores de vidro para linhas de transmissão.” 2005.

A. R. Cantieri, “Método colaborativo para posicionamento de precisão usando vant e vtnt para a inspeção detalhada de torres de distribuição de energia elétrica,” Ph.D. dissertation, Universidade Tecnológica Federal do Paraná, 2020.

L. Jesus and D. Claro, “Detecção semântica de efeitos colaterais na internet das coisas,” in Anais Estendidos do XIV Simpósio Brasileiro de Sistemas de Informação. Porto Alegre, RS, Brasil: SBC, 2018, pp. 105–107.

L. Matikainen, M. Lehtomäki, E. Ahokas, J. Hyyppä, M. Karjalainen, A. Jaakkola, A. Kukko, and T. Heinonen, “Remote sensing methods for power line corridor surveys,” ISPRS Journal of Photogrammetry and Remote Sensing, vol. 119, pp. 10–31, 2016. [Online]. Available: [link].

S. C. G. Junior, “Sistema autônomo para inspeç ões visuais e termográficas em subestaç ões de energia elétrica,” Master’s thesis, Universidade Federal de Minas Gerais, 2017. [Online]. Available: [link].

L. Chen, L. Lin, M. Tian, X. Bian, L. Wang, and Z. Guan, “The ultraviolet detection of corona discharge in power transmission lines,” vol. 5, no. 4B, pp. 1298–1302, 2013. [Online]. Available: [link].

X. Tao, D. Zhang, Z. Wang, X. Liu, H. Zhang, and D. Xu, “Detection of power line insulator defects using aerial images analyzed with convolutional neural networks,” IEEE Transactions on Systems, Man, and Cybernetics: Systems, vol. 50, no. 4, pp. 1486–1498, 2020.

C. Eggert, S. Brehm, A. Winschel, D. Zecha, and R. Lienhart, “A closer look: Small object detection in faster r-cnn,” in 2017 IEEE International Conference on Multimedia and Expo (ICME), 2017, pp. 421–426.

C. Michaelis, B. Mitzkus, R. Geirhos, E. Rusak, O. Bringmann, A. S. Ecker, M. Bethge, and W. Brendel, “Benchmarking robustness in object detection: Autonomous driving when winter is coming,” 2020.