Development Of Monitoring System For Fatigue Specimen Crack Expansion State Based On Machine Vision

In modern industry,fatigue damage is the main cause of the failure of engineering structure and components.Therefore,research on fatigue crack growth of metal materials,predict the fatigue life of engineering structures and components,and ensure that damage failure will not occur during the service life,has the vital significance for practical engineering applications.However,at present the detection of crack length in the fatigue crack growth test in the laboratory mainly uses the microscope method,which has the disadvantages of subjective influence on the test result,labor intensity and long testing time.Therefore,in order to effectively avoid the appeal problem,this paper developed a set of fatigue crack growth expansion state monitoring system based on machine vision.This paper uses some proper hardware such as German Allied Vision Technologies of CCD industrial camera and ADLINK technology industrial computer to developed a set of fatigue crack growth expansion state monitoring system based on machine vision by Qt platform.The system collects the crack image of the test specimen during the test-piece and carries out a series of image processing.After that the monitoring of crack propagation state is realized,and the information related to the test is stored into data base,which is convenient to manage and inquire information.The main research tasks of this paper are as follows:(1)Build the fatigue specimen crack expansion state monitoring system which consists of CCD industrial camera,camera lens,image acquisition card and industrial personal computer to realize the collection of crack image.(2)Utilize the Qt to program procedures,such as the data acquisition,analysis,storage,inquiring and export,which can achieve the functions such as the real-time display of the detection information,and the information storage.(3)The theory of digital image processing on fatigue specimen crack detection based on machine vision is expounded.The collected image is processed with the proper image pretreatment algorithm and get the image.(4)After preprocessing of the image,using the method of combining the Hough transform and region growing to quickly and accurately extract the crack area,and using the improved thinning algorithm to get crack skeleton and the topological structure of the crack.