Fault recognition method and device based on operation images of motor train unit

Abstract

The invention relates to a fault recognition method and device based on operation images of a motor train unit. The method includes the steps that a train historical operation image which has the train model identical with that of a train operation image collected at present is retrieved in a train historical image bank; template matching is performed on a train operation image colleted in real time through a preset train template bank according to the train model in order to achieve alignment and registration of a train and a historical image of a train with the identical model, and the image registered successfully is stored in the train historical image bank; comparison and analysis are performed on the registered train operation image and the train historical operation image with the identical train model, an image area of which the image information is not matched is searched and serves as a fault area, and position information and feature information of the fault area are saved; fault recognition is performed according to the position information and the feature information of the fault area. According to the method, fault recognition on the operating motor train unit is achieved, the efficiency of overhaul operation on the motor train unit is improved, and discovering capacity and warning capacity on hidden faults in the operating process of the motor train unit are improved.

Description

technical field [0001] The invention relates to the technical field of EMU fault identification, in particular to a fault identification method and device based on an EMU running image. Background technique [0002] When the EMU is running at high speed, any small and subtle faults may cause major accidents, so it is very important to enhance the accuracy of EMU fault identification. However, at present, the traditional dynamic image detection system for railway operation failures uses a certain number of operators to manually view and browse each collected image in the form of three shifts to check whether there is a fault in the vehicle. [0003] In the course of realizing the present invention, the inventor finds that there are at least the following problems in the prior art: [0004] The existing method requires a large number of operators, and after a long time of manual viewing, the efficiency is not high, and it is prone to fatigue, which is not conducive to the tim...

AI Technical Summary

Problems solved by technology

[0006] In order to solve the above problems, the present invention proposes a fault identification method and device based on the running images of EMUs, which effectively solves the problems of large workload, low efficiency, and missed detection when only relying on manual browsing to detect image faults

Images