Quality Control Technique Of Mold Repairing By Laser Based On Image Sensor

Modern mold is toward the development of performance and precision,diversification. The consequent processing costs is rising and processing technologyis complex. In order to extend the life of the mold, reduce costs, conserve resources,we adopt the laser with filler wire mold remediation technologies. Laser welding iscompared with the conventional welding method has some advantages, such as lowheat input, small deformation on the work piece, easy to control and so on.In the low-power laser with filler wire welding process, it is difficult to keep theinteraction of the laser beam and the welding wire and the work piece, thecoordination between the wire feed speed and welding speed, because of the smalllaser spot and the fine-wire. Lead to repair efficiency and quality is not stable.According to the filler wire requirements, design a push brushed wire feeder in orderto ensure stability of the laser wire filling welding process. The wire feeder adaptsthe wire feed wheel and straightening device to ensure stiffness of the fine-wire.In order to ensure stability in the low-power laser with filler wire, using coaxialvisual sensing technology to construct coaxial visual collection system. According tothe current source program and make some small changes to the source to edit aVisual C++interface. Collecting the image of the welding process and performingimage processing to obtain the main characteristic information in the welding processare to achieve the prerequisite of stability of the laser with filler wire.In this paper, the coaxial visual image acquisition system is used to collect theimage of the prior to welding wire, weld mold cracks and welding process. As theCCD collected pictures with noise, interference the judgment of laser filler wirewelding stability. This test uses image processing method of denoising, imageenhancement, image binarization, morphological image processing, edge detection,Hough transform to deal with images, and application Krisch operator, Cannyoperator, Hough method to extract the image edge and centerline, to facilitate follow- up and machine processing and calculation. Based on the processing result obtainedby the combination of different image processing methods，get to improve thestability of the wire filling welding seam image processing process and algorithm.Achieve the wire, mold cracks, welds image feature extraction, reduce labor intensity,improve the mold repair efficiency, improve product quality.