Research on Online Identification Method of Foreign Object Invasion in Transmission Lines Based on Image Monitoring

This paper proposes a wire foreign object alarm method for online monitoring of transmission lines to address the issues of low accuracy in monitoring floating objects invading transmission lines and insufficient effectiveness of alarms. This method first constructs an intelligent detection model for floating objects based on a multi-scale RPN network, which realizes the recognition of small-sized hidden danger targets in monitoring scenarios. Subsequently, the article used the improved canny algorithm to extract wires from the images and constructed an effective area for wire foreign object alarm, which was used to remove invalid floating object recognition results such as the ground. Finally, this article conducted tests on the dataset, and the results showed that the proposed method achieved a 97.7% recall rate for wire foreign object alarms, and could greatly reduce false alarms, meeting the needs of intelligent monitoring of transmission lines.