The high-speed railway catenary system is a critical component of railway infrastructure, and any malfunction can cause serious harm. However, due to the complex structure of the catenary system, it is difficult to ensure that all samples in each section will be labeled, which makes it difficult to locate and detect some small sample components. Therefore, this paper proposes a small sample catenary component detection method based on meta-learning. This method can effectively detect unlabeled special segment samples.

High-speed rail, Catenary, Few Samples, Meta-Learning