Localization and Identification of High-Voltage Switchgear Operating Equipment Based on Iterative Segmentation and Topological Matching

High-voltage switchgear in substations can encounter issues like partial discharge during operation, creating a need for automated equipment to minimize the risk to personnel. This paper proposes a detection approach based on segmented iterative analysis, specifically designed for the operational environment of switchgear. The method rapidly processes images of switchgear and identifies the components in operation. By extracting topological data and image features, a feature relationship network is built to represent the operational components. This network is then used to perform semantic extraction of the components through feature matching, facilitating the transition to mechanical automation. Testing with real switchgear data shows that this approach enhances detection accuracy significantly compared to existing methods and exhibits excellent scalability and adaptability.