Existing monitoring methods cannot effectively identify the close crossing and spanning of transmission lines and other facilities (communication cables, trees, buildings, etc.). To meet this challenge, this paper designs an online monitoring system for transmission line crossing and spanning images with intelligent network acceleration services. When transmitting image data, image adaptive compression technology is used to adjust the compression ratio according to the image content, retain the details of the target area while compressing the background part, reduce bandwidth occupancy and improve network efficiency. YOLOv7 (You Only Look Once v7) is introduced to monitor the transmitted transmission line images in real time to identify possible crossing and spanning phenomena. Once a crossing or spanning phenomenon is detected, the system's early warning mechanism is triggered to issue an early warning, thereby ensuring the safety of the transmission line. Through experimental verification, the system designed in this paper shows superiority in monitoring compared with Faster R-CNN, RetinaNet and SSD, with an accuracy of 0.949, a recall rate of 0.913, and an F1 score of 0.931. At the same time, the average response time of the system is 0.496 s, which can quickly identify and respond to the crossing and spanning of transmission lines to ensure timely warning. This study can not only effectively improve the accuracy and real-time performance of transmission line safety monitoring but also solve the problem of image data transmission under bandwidth-limited conditions by optimizing image transmission and compression technology, providing a feasible solution for future intelligent transmission line safety management.