Quantitative Post-Processing Module of Online Automatic Image Inspection for Inner Bubble Defects in a Tire

Abstract

In tire manufacture processing, there are six main procedures including mixing, extruding, calendaring, cutting, building, and inspection. The ripened rubber is wrapping on the steel cord, ply, and different materials in the calendaring processing. When the manufacture processing is completed, the specifications of the tire including profile, dimension, intensity and defect inspection etc. must be verified. The processing of defect inspection is aimed to ensure whether there have been defected inside the tire, to avoid the influence of the durability and safety. Currently, the determinations of defect and its dimension of the tire are detected by defect inspection machine. And the results of the defect inspections can be determined only by the operators, according to self-experience. For this reason, the determination of the tire defect is not a kind of absolute inspection based on the same standard. Therefore, in this investigation, the defect inspection machine and the defect inspection program are combined for the automatic tire defect detection. To verify the Accuracy and the comparison experiments with defect inspection machine have also been carried out. According to the experiment results, the accuracy rate of the proposed measurement system is about 85 % (samples are more than 5000 images) and the maximum deviation between this system and reference system is 5 mm.