Abstract:Ensuring the optimal efficiency of electrical networks requires vigilant surveillance and preventive maintenance. While traditional methods, such as human patrols and helicopter inspections, have been longstanding practices for grid control by electrical power distribution companies, the emergence of Unmanned Aerial Vehicles (UAV) technology offers a more efficient and technologically advanced alternative. The proposed comprehensive pipeline integrates various elements, including preprocessing techniques, deep learning (DL) models, classification algorithms (CA), and the Hough transform, to effectively detect powerlines in intricate aerial images characterized by complex backgrounds. The pipeline begins with Canny edge detection, progresses through morphological reconstruction using Otsu thresholding, and concludes with the development of the RsurgeNet model. This versatile model performs binary classification and feature extraction for power line identification. The Hough transform is employed to extract semantic powerlines from intricate backgrounds. Comparative assessments against three existing architectures and classification algorithms highlight the superior performance of RsurgeNet. Experimental results on the VL-IR dataset, encompassing both visible light (VL) and infrared light (IR) images validate the effectiveness of the proposed approach. RsurgeNet demonstrates reduced computational requirements, achieving heightened accuracy and precision. This contribution significantly enhances the field of electrical network maintenance and surveillance, providing an efficient and precise solution for power line detection.

What problem does this paper attempt to address?

The problem that this paper attempts to solve is: how to detect transmission lines efficiently and accurately from aerial images. Although traditional methods such as manual patrol and helicopter inspection have been widely used, they are less efficient and have potential dangers. With the development of unmanned aerial vehicle (UAV) technology, a more efficient and advanced alternative is provided. Specifically, the paper aims to solve the problem through the following points: 1. **High - precision detection in complex backgrounds**: Achieve high - precision detection of transmission lines in complex aerial image backgrounds. 2. **Reduce computational requirements**: Develop a method with lower computational requirements to improve the efficiency and accuracy of detection. 3. **Adapt to different lighting conditions and terrains**: Ensure that the method can maintain good performance under different lighting conditions and terrains. 4. **Automation and intelligence**: Utilize deep - learning models and classification algorithms to realize the automation and intelligence of transmission line detection. To this end, the paper proposes a comprehensive pipeline that combines multiple techniques, including pre - processing techniques, deep - learning models (DL), classification algorithms (CA) and Hough transform, to effectively detect transmission lines from complex aerial images. ### Key techniques and methods - **Canny edge detection**: Used to identify edge features in images. - **Morphological reconstruction**: Use Otsu thresholding for image enhancement. - **RsurgeNet model**: A binary - classification model used to distinguish the presence or absence of transmission lines and extract meaningful features. - **Hough transform**: Used to extract semantic transmission lines from complex backgrounds. ### Experimental verification The experimental results show that the RsurgeNet model performs better than the existing three architectures and classification algorithms on the VL - IR dataset, demonstrating higher accuracy and precision. In addition, this model also has lower computational requirements, further improving its feasibility and efficiency in practical applications. ### Main contributions - Proposed a comprehensive transmission line detection pipeline that can effectively deal with complex backgrounds and variable environments. - Developed the RsurgeNet model to achieve efficient transmission line detection and feature extraction. - Conducted extensive experimental verification on visible - light and infrared images, proving the effectiveness of this method. In general, this research provides an efficient and accurate solution for the maintenance and monitoring of power networks, significantly improving the technical level of transmission line detection.

Overhead power line detection from aerial images using segmentation approaches

Overhead Power Line Detection from UAV Video Images

Powerline extraction from aerial and mobile LiDAR data using deep learning

Power Line-Guided Automatic Electric Transmission Line Inspection System

Aerial Inspection of High-Voltage Power Lines Using YOLOv8 Real-Time Object Detector

End-to-End Powerline Detection Based on Images from UAVs

DUFormer: Solving Power Line Detection Task in Aerial Images using Semantic Segmentation

Intelligent Detection of Vegetation Encroachment of Power Lines with Advanced Stereovision

Electric power line patrol operation based on vision and laser SLAM fusion perception

Fault Detection in Power Equipment via an Unmanned Aerial System Using Multi Modal Data

Enhancing Power Lines Detection Using Deep Learning and Feature-Level Fusion of Infrared and Visible Light Images

Automatic Insulator Detection for Power Line Using Aerial Images Powered by Convolutional Neural Networks

Autonomous Point Cloud Segmentation for Power Lines Inspection in Smart Grid

Cable Fault Detection and Extraction Based on UAV Collected Images

Automatic Power Line Detection for Low-Altitude Aircraft Safety Based on Deep Learning

Detection of Potential Faults in the Electricity Distribution Network Using Unmanned Aerial Vehicles and Thermal Cameras Through Deep Learning Methods

Image Detection Technology on Pin Defect to Overhead Power Lines

Power Line Detection Based on Maxtree and Graph Signal Processing

UAV Visual and Thermographic Power Line Detection Using Deep Learning

Online Monitoring of Overhead Power Lines Against Tree Intrusion Via a Low-cost Camera and Mobile Edge Computing Approach

High Speed Automatic Power Line Detection and Tracking for a UAV-Based Inspection