The subway has the advantage of convenience and speed in our daily life, which brings convenience to our traffic. At the same time, the high load operation behind this has brought huge challenges and pressures to safety issues. In view of track safety issues, people have proposed some detection methods, such as fasteners defect detection, rail crack detection, sleeper broken detection, sensor board shift detection, and foreign object detection. Vision-based recognition methods such as Haar-Like + Adboost, HOG + SVM, and the use of laser sensors are often not particularly ideal. Therefore, improvements have been made on the basis of yolov3 target detection algorithm, and a multi-level feature map cascade based on feature pyramid is proposed. The Densely connected feature map performs the task of identifying each sub-target of the defect, and finally realizes the recognition of high precision and frame rate on the 3080Ti deep learning platform.
Track Defect Recognition Algorithm Based on Deep Learning
Lect. Notes Electrical Eng.
International Conference on Electrical and Information Technologies for Rail Transportation ; 2021 October 21, 2021 - October 23, 2021
Proceedings of the 5th International Conference on Electrical Engineering and Information Technologies for Rail Transportation (EITRT) 2021 ; Chapter : 45 ; 397-408
2022-02-23
12 pages
Article/Chapter (Book)
Electronic Resource
English
Track Defect Recognition Algorithm Based on Deep Learning
British Library Conference Proceedings | 2022
|Track Defect Recognition Algorithm Based on Deep Learning
TIBKAT | 2022
|Deep Learning-Based Image Analysis to Track Vehicles
IEEE | 2024
|