Train anomaly recognition detection method and system

Abstract

The invention discloses a train anomaly recognition detection method and system and belongs to the technical field of automatic image recognition. The recognition method includes the steps of collecting whole train depth information of a train to be detected and a reference train, matching a train model and a train number with the whole train depth information of the train to be detected and the reference train for recognizing faulty components, and matching depth information of the faulty components with that of reference parts for determining faulty parts. The recognition system comprises a collection module for collecting the whole train depth information of the train to be detected, a reference train information acquisition module and a faulty part detection module. A method of collecting three-dimensional image information of a running train with the depth information overcomes the visual function limitation of traditional two-dimensional image information so as to facilitate quick find of actual faulty points of a detection worker, reduce the false positive rate of the system and improve the maintenance efficiency of the detection worker.

Description

technical field [0001] The invention discloses a detection method and system for identifying train anomalies, belonging to the technical field of automatic image recognition. Background technique [0002] The traditional abnormality detection of the whole train parts is usually manually detected. This method needs to be detected when the train enters the station or enters the warehouse. Since the train (including freight cars, passenger cars, EMUs and other types of rail trains) stops for a long time Coupled with the fact that there are many parts and complex structures on the train, it is difficult for maintenance personnel to pay attention to the normal status of each part in a short period of time. In particular, there are occlusions between the parts at the bottom and top of the train and there are visual blind spots, which not only further increases the probability of missed detection of abnormal train parts, but also further reduces the efficiency and accuracy of abnor...

AI Technical Summary

Problems solved by technology

[0002] The traditional abnormality detection of the whole train parts is usually manually detected. This method needs to be detected when the train enters the station or enters the warehouse. Because the train (including freight cars, passenger cars, EMUs and other types of rail trains) stops for a long time In addition, there are many parts and complex structures on the train, it is difficult for maintenance personnel to pay attention to the normal status of each part in a short period of time

In particular, there are occlusions between the parts at the bottom and top of the train and there are visual blind spots, which not only further increases the probability of missed detection of abnormal train parts, but also further reduces the efficiency and accuracy of abnormal detection

[0003] In addition, the train is in the running state most of the time. During the operation of the rail train, due to the impact of debris, relative movement, impact, vibration, etc., the parts will be worn, abnormally loose and other problems. When the train enters the station or completes the operation Repairs are carried out after the number of kilometers has been put into storage, which will cause abnormalities in the parts at the bottom of the train to not be discovered and repaired in time, and will eventually cause great safety hazards to the operation of rail transit

Images