Images collected by linear scan cameras are stretched and compressed due to the speed regulation of trains. This condition changes the shape of objects and considerably increases the false and missed alarm rates. Moreover, small size and dense distribution of the key components in railway trains increase the difficulty in fault diagnosis. Therefore, a two-module troubleshooting and positioning methodology is proposed in this study for state diagnosis of key small components of high-speed running trains. Firstly, the image with deformation is reshaped to be exactly the same as the standard one via omnidirectional scale correlation normalization (OSCN), which performs well even in low-texture, high-light situations. Secondly, we propose three feature-enhanced models to expand the receptive field of deep feature maps. This novel detector, namely, refine-inception net (RIN), not only reduces the rate of missed and false detection, but also minimizes the influence of target size and occlusion. Augmentation is performed to increase robustness because the detector is data driven. Experimental results show that the optimal strategy combining OSCN and RIN can troubleshoot high-speed trains of different models with an accuracy higher than 99%. Our method can be extended to foreign object recognition on the train roof while maintaining railway safety.

Pantograph slider is a critical component of the electrical equipment to obtain power sources for the electric locomotive, whereas the interaction between the slider and the overhead contact line can cause persistent surface wear of the collector strip. The condition monitoring for the level of abrasion of the pantograph slider is significant to avoid major damage. An innovative approach based on multi-view analysis is proposed for monitoring the surface wear of the slider in a running train, accurately measuring the residual thickness of the slider and locating the wear defects appearing. Specially, the three-dimension (3D) reconstruction technology is used for detecting the wear quantity between the upper and lower edges of the slider. First, target edges of the slider are determined using the proposed subpixel edge detection algorithm. Second, mutual matching of the edge points between the multi-view images is finished using the epipolar geometric constraint. Moreover, the 3D reconstructions of the matching points are computed with the principle of triangulation. Finally, the wear volume of the slider can be determined. Experimental results show that the proposed method can successfully measure the abrasion of the pantograph slider under complex conditions, which demonstrate the applicability, effectiveness, and good performance of the proposed approach.

A good on-line contact, characterized by high-precision measurement, between the pantograph and catenary (PAC) ensures the safety of PAC and subsequently high-speed train operation. Currently, image processing is the mainstream measurement method for detecting contact points, in which the contact point's real-time tracking in the image sequences with complex background is the method's core difficulty. In this paper, a contact point surveillant method that can run during the actual operation of the railway is proposed. This surveillant method can be initialized by the global robust detection method and can do real-time tracking by means of kernelized correlation filters, which has high stability in regional tracking. With the help of result refinement and result confidential evaluation methods, this surveillant method can stably and independently obtain real-time and accurate results, which further gives the state analysis based on 3-D positions. The proposed method is tested on image data sets collected from a high-speed train running at 200 km/h, and results corroborate that a high-speed, high-precision, and stable contact point tracking is achievable, which ensures the state measurement's feasibility by means of the contact point trajectory.

Online autonomous perception of pantograph catenary system status is of great significance for railway autonomous operation and maintenance (RIOM). Image sensors combined with an image processing algorithm can realize the automatic acquisition of the pantograph catenary condition; however, it is difficult to meet the demand of long-term stable condition acquisition, which restricts the implementation of online contact state feedback and the realization of railway automation. This paper proposes an online intelligent perception of the pantograph and catenary system (PCS) status based on parameter adaptation to realize fast and stable state analysis when the train is in long-term operation outdoors. First, according to the feature of the contact point, we used histogram of gradient (HoG) features and one-dimensional signal combined with a KCF tracker as the baseline method. Then, a result discriminator located by L1 and hash similarity constraints was used to construct a closed-loop parameter adaptive localization framework, which retrieves and updates parameters when tracking failure occurs. After that, a pruned RefineDet method was used to detect pantograph horns and sparks, which, together with the contact points localization method, ensure the long-term stability of feature localization in PCS images. Then, based on the stereo cameras model, the three-dimensional trajectory of the whole pantograph body can be reconstructed by the image features, and we obtained pantograph catenary contact parameters including the pantograph slide posture, contact line offset, arc detection, separation detection, etc. Our method has been tested on more than 16,000 collected image pairs and the results show that the proposed method has a better positioning effect than the state-of-art method, and realizes the online acquisition of pantograph catenary contact state, representing a significant contribution to RIOM.

A safe train operation relies on the well-contact of pantograph and the power grid above, therefore identifying the state of pantograph plays an vital role. Among all of the malfunctions, pantograph offset is a strong reflection of the state. We have proposed a new approach to reconstructing the three-dimensional (3D) information of the bow by substituting the offset with connection of left and right horns. To locate the region of pantograph horn, we refer to an efficient deep learning method, named Single Shot MultiBox Detector (SSD). In the located area, region growing or Wiener filtering is applied to extract connected components and enhance the prospects. For processed images, grayscale morphological gradients is adopted to obtain image edges, on which Harris corner detection can provide dozens of potential corner-points. These points containing one correct horn point needs selection by assuming the lowest one is the best, only when the background noise is as low as possible. After attaining the two-dimensional (2D) image coordinates of the horns, binocular stereo vision method makes contribution to reconstructing 3D coordinates. Using the 3D coordinate line of the left and right horns to represent the pantograph offset can easily reflect the degree of deviation by comparing it with the initial location of pantograph. Our approach is of great significance to prevent malfunction, which are about to arise later, such as horn loss or deflection, spark of pantograph and catenary system contact-point. The detection of the pantograph offset provides a strong guarantee for maintaining railway traffic safety.