With the rapid development of deep learning technologies, researchers have begun to utilize convolutional neural network (CNN)-based object detection methods to detect multiple catenary support components (CSCs). The literature has focused on the detection of specified large-scale CSCs. Additionally, CNN architectures have faced difficulties in identifying overlapping CSCs, especially small-scale components. In this paper, a unified CNN architecture is proposed for detecting all components at various scales of CSCs. First, a detection network for CSCs with large scales is proposed by optimizing and improving Faster R-CNN. Next, a cascade network for the detection of CSCs with small scales is proposed and is integrated into the detection network for CSCs with large scales to construct the unified network architecture. The experimental results demonstrate that the detection accuracy of the proposed CNN architecture can reach 92.8%; hence, it outperforms the popular CNN architectures.@en

Brace Sleeve (BS) plays an essential role in connecting and fixing cantilevers of railway catenary systems. It needs to be monitored to ensure the safety of railway operations. In the literature, image processing techniques that can localize BSs from inspection images are proposed. However, the boxes produced by existing methods can contain incomplete and/or irrelevant information of the localized BS. This reduces the accuracy of BS condition diagnosis in further analyses. To address this issue, this paper proposes the use of an action-driven reinforcement learning method that adopts the coarse-localized box provided by existing methods, and finds the movements needed for the box to approach to the true BS position automatically and accurately. In contrast to the existing methods that predict one position of the box containing a BS, the proposed action-driven method sees the localization problem as a dynamic position searching process. The localization of BS is achieved by following a sequence of actions, which in this paper are position-moving (up, down, left or right), scale-changing (scale up or scale down) and shape-changing (fatter or taller). The policy of selecting dynamic actions is obtained by reinforcement learning. In the experiment, the proposed method is tested with real-life images taken from a high-speed line in China. The results show that our method can effectively improve the localization accuracy for 81.8% of the analyzed images. We also analyze cases where the method did not improve the localization and suggest further research lines.@en

The goal of this paper is to evaluate from a multi-objective perspective the performance on the detection of catenary support components when using state-of-the-art deep convolutional neural networks (DCNNs). The detection of components is the first step towards a complete automatized monitoring system that will provide actual information about defects in the catenary support devices. A series of experiments in an unified test environment for detection of components are performed using Faster-CNN, R-FCN, SSD, and YOLOv2. Through the comparison of different assessment indicators, such as precision, recall, average precision and mean average precision, the detection performance of the different DCNNs methods for the components of the catenary support devices is analyzed, discussed and evaluated. The experiment results show that among all considered methods, R-FCN is the more suitable for the detection of catenary support components.@en