Online Rail Fastener Detection Based on YOLO Network

Abstract

Traveling by high-speed rail and railway transportation have become an important part of people's life and social production. Track is the basic equipment of railway transportation, and its performance directly affects the service lifetime of railway lines and vehicles. The anomaly detection of rail fasteners is in a priority, while the traditional manual method is extremely inefficient and dangerous to workers. Therefore, this paper introduces efficient computer vision into the railway detection system not only to locate the normal fasteners, but also to recognize the fasteners states. To be more specific, this paper mainly studies the rail fastener detection based on improved You can Only Look Once version 5 (YOLOv5) network, and completes the real-time classification of fastener states. The improved YOLOv5 network proposed contains five sections, which are Input, Backbone, Neck, Head Detector and a read-only Few-shot Example Learning module. The main purpose of this project is to improve the detection precision and shorten the detection time. Ultimately, the rail fastener detection system proposed in this paper is confirmed to be superior to other advanced algorithms. This model achieves on-line fastener detection by completing the "sampling-detection-recognition-warning" cycle of a single sample before the next image is sampled. Specifically, the mean average precision of model reaches 94.6%. And the model proposed reaches the speed of 12 ms per image in the deployment environment of NVIDIA GTX1080Ti GPU.

An improved YOLOV5-based algorithm for detecting unauthorized personnel intrusion on tracks for reducing railway security risks

“…With the emergence and development of YOLO, it is increasingly used in the field of defect detection [4] [5].In July 2021 W Liu et al designed a real-time Yolov5-based railroad signal light detection system using an improvement based on Hof circle variation [6].In 2022 J Li et al proposed an improved YOLOv5 network consisting of five parts, i.e., input, trunk, neck, head detector and read-only lens less example learning module to improve the detection accuracy and shorten the detection time [7].2022 Fan Zhang et al proposed a SIP-YOLOv5 network to detect negative obstacles, especially by adding a small target detection layer and improved coordinate attention in YOLOv5 to improve the accuracy [8].2022 Z Su et al proposed a track fastener defect detection method based on improved YOLOv5, which utilizes the K-means algorithm to analyze the size of the target frame of fastener defects and determine the size of the most preferred frame. Secondly, the attention mechanism is combined with multiscale fusion thereby increasing the speed and accuracy [9].Yolo belongs to the single-stage model which is more advantageous in terms of detection speed.…”

FPGA Implementation of Extended Kalman Filter for Parameters Estimation of Railway Wheelset

“…However, the creep curves for different types of contaminated rails not clear enough, so systematic work on the influence of creep curves for all conditions is required. A lot of research is being carried out for the estimation of wheel-track contact conditions mainly scholars use model-based techniques [12][13][14][15].…”

Few-shot pump anomaly detection via Diff-WRN-based model-agnostic meta-learning strategy

“…[6][7][8][9][10][11] However, considering the operating environment is usually accompanied by high temperature and enormous pressure, health monitoring is always prohibitively expensive or hazardous. 12,13 Few-shot learning (FSL) provides the basis for developing an effective classifier in data-scarce situations. [14][15][16] In this scenario, numerous solutions have been explored to prevent performance degradation.…”

“…6–11 However, considering the operating environment is usually accompanied by high temperature and enormous pressure, health monitoring is always prohibitively expensive or hazardous. 12,13…”