Research on Remote Sensing Map-Based Power Transmission Tower Positioning System Based on Deep Learning

During the continuous expansion of the transmission network, the transmission lines are inspected by manual inspection and unmanned aircraft inspection. However, all of them may fail in the environment where disasters occur, resulting in failure to identify and locate grid faults in a timely manner. In this paper, we propose an intelligent system for transmission tower identification and location based on deep learning algorithm and remote sensing technology. The proposed technique can locate transmission towers, update transmission line data, and discover potential external breakage under extreme conditions. The remote sensing image transmission pole tower localization system is optimized by exploiting YOLOv5l model. Super-resolution module and overlap slicing layer are developed to improve the accuracy rate of identification and localization. In this way, the processing of the original remote sensing image data is significantly improved.Furthermore, a coordinate mapping algorithm is established to realize the geographic coordinate positioning of transmission tower equipment on remote sensing images. In addition, a dataset of transmission tower images from remote sensing satellite (TIRS) is constructed to verify the effectiveness of the system. The YOLOv5l model is trained on the TIRS dataset and tested on remote sensing maps. The results have high accuracy and support multiple target localization on large size remote sensing images. The localization speed for a single device is in millisecond level, which can greatly reduce the time cost and labor cost of transmission line data update.