The Algorithm Research And Software Design Of Surface Defect Inspecting System Based On Machine Vision

In order to continually raise the speed and efficiency of production, the automatic and on-line inspecting technology about surface defect of metal production becomes more and more important. This paper designs the automatic system, which can inspect surface defect of metal production timely and un-scathing, researching on many defects on the surface of metal production.The system using black and white CCD and multi-channel vision DAQ card as vision acquiring part raises the speed of inspecting system and induces the capability requirement of CCD, which can realize timely and online inspecting easily under the current condition. This system can select threshold automatically and pick up the useful information of production from the image. By using the pickup information, it can scan the whole image and get the production's position and dimension information, realizing the system's automatic measuring. In virtue of these thresholds and the production's information, the system can divide the image and get the information of defect, which provides for the latter defect recognition.The system looks on the theory of computer vision as inspecting theory. In terms of current technology and error theory and using module design idea, I have done a lot of experiment and much arithmetic simulation of image processing and designed out the algorithm suiting for the product surface defect inspecting on the base of image processing theory at present. What I have done is as followed:1. Research on the image acquiring and saving methods on computer. Assign computer EMS memory and save two channels'area CCD data in it, helping for the next image processing by wiping off jam.2. Design out the different thresholds selecting algorithm for image division. By the selecting thresholds, accomplish picking up the defect area automatically.3. On account of the characteristic of all kinds of defect, I have done much research on such theory as image smooth, margin sharpen, area divide etc. In the end, inspecting algorithm is designed out adapting to this system and simulated on Matlab6.5 successfully.4. Accomplish the software programming and translate the successfully simulated inspecting algorithm into the programming environment of VC6.0, consummating the software of the system.