The Machine Vision Measurement Of The Round Steel’s Outline And Surface Crack

As the rapid development of modern industry，the demands of various types ofsteel in various industrial fields increase sharply. As one of the common steal materials,round steel, it’s diameter and surface quality are an important indicator to measure theround steel’s quality. So the significance of quality controlment in the process of itsproduction has become increasingly significant，especially how to realize onlinedetection has become the primary concernment of the major manufacturers.The crackaffects the normal use of the product，and the diameter parameter affects where theround steel can use.The machine vision round steel’s outline and crack measurement system is basedon the framework of the computer vision theory. It capture digital image by using CCDcamera，then using Visual C++edit programs to process encoded gray scale bitmap，soas to achieve the purpose of detection. This system can realize non-contact onlinemeasurement.As one of the functions of the system in this subject, outline measurement，it iscontrolled by computer to light line laser through the computer’s serial. By lightinground steel，at the same time，the CCD camera captures images. On the acquiredimages，it needs a series of operations，binaryzation，calibration，image expansion，image median filtering，refinement，ellipse fitting and so on. So it can restore the nearlycircle outline，and then obtain the diameter，radius and other parameter.Crack detection is another function of the system in this topic. Crack detectiondoes not require laser lighting. Computer controls line laser going off through the serialport，while the CCD camera captures images. Then doing some operations, such asimage pre-processing, image edge detection,binaryzation, calibration, to separate thecrack information, thus calculate the linear crack’s width and length, point crack’s areaand other parameters.The hardwares of the system include a CCD camera, laser and a computer. Usinglaser illumination, the CCD camera capture images. Using the computer to process，in order to achieve the sensory function of the visual observation and brain analysis，andidentify the size of the outline, the information of crack and so on. In which theillumination of the laser is not necessary.