Data analysis in visual power line inspection: An in-depth review of deep learning for component detection and fault diagnosis

摘要

The widespread popularity of unmanned aerial vehicles enables an immense amount of power line inspection data to be collected. It is an urgent issue to employ massive data especially the visible images to maintain the reliability, safety, and sustainability of power transmission. To date, substantial works have been conducted on the data analysis for power line inspection. With the aim of providing a comprehensive overview for researchers interested in developing a deep-learning-based analysis system for power line inspection data, this paper conducts a thorough review of the current literature and identifies the challenges for future study. Following the typical procedure of data analysis in power line inspection, current works in this area are categorized into component detection and fault diagnosis. For each aspect, the techniques and methodologies adopted in the literature are summarized. Valuable information is also included such as data description and method performance. In particular, an in-depth discussion of existing deep-learning-based analysis methods of power line inspection data is proposed. To conclude the paper, several study trends for the future in this area are presented including data quality problems, small object detection, embedded application, and evaluation baseline.