Research On Rapid Sorting Method Of Refurbished Tires Based On Machine Vision

In recent years,with the increase of China's automobile demand,as the automotive upstream industry of the tire industry,the demand is also soaring,at the same time,the number of waste tires are also increasing.From the perspective of resource recycling,it is a very good choice to refurbish used tires.However,China's tire refurbishment is relatively low,mainly because of quality and safety issues.Therefore,it is imperative to realize the automatic screening system for refurbishing used tires to improve the detection quality.Based on the research on the rapid sorting method of refurbished tires based on machine vision,this paper designs a set of detection system of refurbished tires.The specific research contents are as follows:(1)This paper introduces the current situation of tire refurbishment industry.The development status of machine vision technology at home and abroad is analyzed.(2)According to the requirements of car Tyre refurbishment(GB14646-2007),the research objectives were clarified,the research scheme was drawn up,and the testing system was designed and analyzed.(3)The method of template matching and spatial coordinate calculation is used to locate the characters on the side of the tire.Thecharacters after the positioning are segmented and classified by the trained classifier to achieve the effect of character recognition.(4)The internal defects of waste tires were analyzed and the image algorithm for defect recognition and detection was designed.The experiment was verified from the two directions of target detection and semantic segmentation,and the training analysis of different types of samples was carried out to select the appropriate training sample type.Three different types of network models were used for training comparison,and a variety of pattern types of tires were used to verify the generalization ability of the refurbished tire test model.(5)Through software Halcon and C# joint programming,design simple human-computer interaction interface.The hardware layout of the system is analyzed to build a complete detection system.