State Detection of U-Shaped Hoop in Railway Catenary Based on Semantic Segmentation

Manual detection of the U-shaped hoop state of the railway catenary is ineffective in accuracy. Some automatic methods have slow speed and low accuracy. In order to solve this problem, a U-shaped hoop state detection method based on semantic segmentation is proposed. Firstly, the bolt and nut area of the U-shaped hoop is segmented by the semantic segmentation model PSPNet, and then the state of the U-shaped hoop is judged according to the area ratio of the segmented area. On this basis, the lightweight network Mobilenetv2 is used as the backbone network to accelerate the segmentation speed, and the CBAM attention mechanism and the FPN feature pyramid structure are integrated to improve the segmentation accuracy. The experimental results show that the improved model in this paper is superior to the original model in terms of speed and accuracy, and at the same time satisfies the requirement of using the segmentation results to judge the state of the U-shaped hoop.