Rail fastener disease detection method and system

Abstract

The invention provides a rail fastener disease detection method and system. The method comprises steps that images of to-be-detected rail fasteners are acquired; the images of the to-be-detected railfasteners are inputted to a trained convolutional neural network, and the original output result of the convolutional neural network is obtained and outputted; based on the preset decision discrimination conditions, the original output result is detected and discriminated, and the images of the target rail fasteners having diseases are determined. The method is advantaged in that the powerful adaptive learning capability of the convolutional neural network is utilized, disease characteristics of the fasteners are automatically learned, the function of automatically detecting the fastener diseases is realized, the detection speed is fast, and detection is accurate.

Description

technical field [0001] The embodiments of the present invention relate to the technical field of rail transit safety, in particular to a method and system for detecting defects of rail fasteners. Background technique [0002] At present, with the rapid development of road traffic, public transportation is shifting from the construction period to the maintenance period. As the basis of railway operation, ensuring the safety and reliability of railway tracks is an important task for daily maintenance of railways. Rail fasteners are intermediate parts connecting rails and sleepers. In the operation of railways, fastener installation errors, fastener loss, tripping and other defects will occur, which will cause loose rails, affect the normal passage of railways, and even cause safety accidents. How to timely and accurately discover the location of the disease is an important task to ensure the safe operation of the railway. [0003] In the prior art, manual visual inspection of...

AI Technical Summary

Problems solved by technology

[0003] In the prior art, the fasteners are often inspected manually with naked eyes, and this detection method will consume a lot of manpower and time

There is also a method of detection through computer vision, such as using gray-scale path to Canny edge detection to judge the location of fasteners, but this method has poor detection effect and low robustness

Images