Research on Robot Vision Detection Method for Scratch Defects of Flat Glass Based on Area Array CCD

With the increasingly wide application of glass in aviation, automobile and other fields, high-quality glass has a vast demand market. Along with the popularization of robot automation production lines, the glass defect intelligent detection system needs to be developed urgently. This paper takes the detection of flat glass scratch defects as the research object, a detection experiment platform is designed and constructed based on area array CCD industrial camera, and a vote filtering algorithm is proposed, which can effectively remove the discrete noise points of glass scratch defect. The method of extracting the length and width of the feature parameters of glass scratch defect is proposed by applying the theory of minimum enclosing rectangle. The experimental results show that the constructed experimental hardware platform and the proposed detection methods can automatically, real-timely and accurately achieve scratch defect analysis and glass qualification. It has important robot engineering application value and significance.