Study On Detection Algorithm Of Steel Cord Conveyor Belt Monitoring System Based On X-ray

Steel cord conveyor belt,as the “big artery” of transport in the field of heavy industry,will determine the production and transport capacity of enterprises.It will improve the production efficiency that if enterprises can use it reasonably and effectively.On the contrary,if enterprises can't guarantee the safe operation of the conveyor belt,the resulting major accident will not only affect the production,and even let coal miners pay their precious life.Therefore,combined with computer and image processing and other professional technical,developing a set of efficient,real-time,accurate steel cord conveyor belt detection system to achieve intuitive,high-speed on-line detection and accurate off-line analysis,accurately positioning conveyor belt defects or faults,qualitatively and quantitatively analyzing them,make an important contribution for the safety and long-term development of the coal industry.In the thesis,the on-line detection and off-line analysis algorithm of the steel core conveyor belt monitoring system based on X-ray has been studied and explored in detail.According to the periodic texture characteristics of the conveyor belt image,a linear texture coding algorithm based on LBP is proposed to detect defects and identify joints on-line.And the detected joint images,defect images and all the conveyor belt images collected by the linear array detector are stored in the user specified path to facilitate other functional modules of the monitoring system analyze and process the conveyor belt image data more deeply.In the research process of the off-line analysis algorithm,the Gaussian blur and the USM algorithm are used to optimize the conveyor image,so that the contrast of the gray scale mutation is effectively enhanced,highlighting important details,and the conveyor belt image with noise is made clearer.Based on the careful analysis of the detected defect images and joint images,an off-line analysis algorithm based on Harris corner detection principle and Hough transform linear detection principle is proposed to analyze the defect images and joint images accurately.The off-line analysis algorithm can generate a PDF report that contains user information,conveyor information,defect information,and joint mark.It can be printed at any time to easily view the conveyor belt status or to perform accurate maintenance of the conveyor belt.In the software development process of the monitoring system,Microsoft Visual C++ 6.0 is used as the software development platform.Based on C++ language,adopting object-oriented programming,combines multithreading technology,database technology and open source library(MFC,CxImage,OpenCV etc.),the successful realization of the system software development.And the monitoring system has been applied to a number of coal mines in Shanxi Province,after testing,fully meet the actual production requirements of coal mines.According to the feedback information during the operation of the conveyor belt,and constantly optimize the function of the system.