Design And Research Of Railway Tunnel Vault Disease Detection System Based On Machine Vision

With the expansion of railway construction,the quantity of tunnel is growing rapidly.But the detection of tunnel disease has been always a technical problem in the procedure of maintenance.As the application of automatic equipments is in its infancy,artificial reading and recording is still the main method in tunnel testing in our country at present.The high efficiency and accurate precision of disease detecting technology based on machine vision which has a broad application prospect can make up for the deficiency of traditional testing methods.Based on the in-depth study of disease detecting technology at home and abroad,this thesis designed a set of railway tunnel vault disease detecting system based on machine vision.This paper first analyzed the principle and application of a variety of NDT methods and determined the technical route of machine vision detecting system by comparing their characteristics and applicable range.Then the design of the system and working process was completed under the overall consideration of characteristics and detection requirements of railway tunnel.And then the detecting system which is consisted with hardware and software was buildup according to related technical indexes.After that,this paper first accomplished the simulation measurement of width by the use of stripe images.And then the field experiments of tunnel disease detecting was carried out and the detection performance of the system was analyzed according to the results.The hardware of the detecting system mainly includes power supply,industrial personal computer,electric control box,linear CCD camera,image capture card,the LED light source and GPS sensor module.The software was developed by the object-oriented C++ programming language under the environment of Microsoft Visual Studio 2008.The main functions of the software are image acquisition,video display,equipment control and disease information measurement,etc.Thanks to its high signal-to-noise ratio,high positioning accuracy and low probability of edge response,Canny edge detection algorithm was adopted in crack identification and positioning.This thesis not only designed a set of railway tunnel vault disease detecting system,but also introduced the system's working principle and the process of design.It provides a new solution for the tunnel diseases detection and will be of great significance in ensuring the safety of railway operation.