Feature Extraction Of Laser Mold Repair Dased On Acoustic Emission Aided Coaxial Vision

Laser mold repair is a complex process. At this stage, most of the repair for mold repair using manual methods. In this paper, in-depth research the mechanism of laser mold remediation for the beneficially exploring to the repairing of mold by laser automatically and intelligently.Designed coaxial laser mold repairing system, successfully solved the problem that laser optical path consist with the CCD optical path to separate them and independently working. Designed rangefinder CCD defocus detection system; designed acoustic signal acquisition system.Analyze the coaxial visual sensing mold cracks pictures,successfully extract the centerline of crack picture. Lay the foundation of mold crack track. Use coaxial vision from the pool just above the weld pool observe the weld pool, extract the feature information of weld pool picture. Have a better understanding on visual sense of quality control in the mold repairing. Using a microphone sensor collect the acoustic signal producing in the mold repairing process, the acoustic signals were wavelet transform and time-domain analysis, discussed the characteristics of the information under different conditions of acoustic signals, make that acoustic signal in the mold repair quality control has been further theoretical foundation. Finally, using the lateral CCD vision sensing system detect the amount of defocus to determine whether splash producing; RMS as anther characteristic quantity based on the researching for different conditions. Achieving detecting in many aspects to laser mold repair using vision sensing and acoustic signal sensing; build the foundation of precise control of the quality of the mold repair and muti-information merging in laser mold repair.