Research On Tire Gluing Quality Inspection System Based On Machine Vision

As one of the most practical technologies in industrial inspection technology,machine vision inspection technology will definitely shine in the detection of industrial products in the future.The apex is one of the important components of automobile tire inner tubes,and its quality is directly related to its quality.Whether the multiple performance indicators of the car meet the standard.At present,delta glue manufacturers mostly adopt traditional manual detection methods in the production process.This method has the problems of high labor cost,high false detection rate,and low efficiency,which greatly affects the production efficiency of the enterprise,and it is not in line with"Made in China 2025" The guiding ideology of intelligent manufacturing,therefore,this paper designs a set of tire apex quality inspection system based on machine vision technology based on the actual production situation.The main content and work results are as follows:(1)The visual inspection system is designed in combination with the existinggluing and squeezing method.The comprehensive analysis shows that the visual module is most suitable for the laser triangulation method.The measurement model uses a laser generator perpendicular to the measured surface and is integrated with the visual sensor.40° included angle method,this method can reduce the error caused by the laser tilt.The system’s software and communication modules are developed and implemented based on the.NET Framework and OPC technology,and are characterized by convenience,speed,and robustness.(2)Analyze the basic principles of camera calibration,and propose a zigzag calibration method based on Zhang Youzheng’s calibration ruler.That is,the 10.0mm sawtooth calibration ruler is used to complete the coordinate calibration of the visual sensor system,and the calibration parameters of the visual sensor module are obtained through experiments.It is concluded that the model resolution is 0.0284mm/pixel in the Z direction and 0.0455mm/pixel in the X direction.(3)Analyze and analyze the processing flow of the glulam image through experiments,improve the collection efficiency of the image and the processing speed of the algorithm by setting different regions of interest,use the median filtering method to clean up the clutter in the image,and use the histogram to obtain Glue binary image,using Otsu method to achieve important feature acquisition,using Canny operator to detect and identify the edge of the gluing area of gluing,based on image morphology to achieve the highlighting of image features,all algorithms are designed using C#.NET tools achieve.(4)This paper summarizes the four types of defects of Glue and formulates different detection strategies,and proposes a feature extraction and recognition method for the problem of edge dislocation and excessive overlap.Complete the hardware selection and installation based on the existing delta glue production indicators and the accuracy requirements of the system measurement.The visual inspection module adopts area array CMOS sensor and 500mV,650nm semiconductor single-line laser generator.The software module runs on the industrial computer.The system is designed with a safe and concise software interface to enhance the best user experience and provide good reuse.With the extensible initial parameter preset template,the output result is displayed in XML data format.After the implementation of the quality inspection method,the detection and comparison experiments of 100 samples showed that the system’s missed detection rate was 0%,the false detection rate was 3%,and the detection efficiency was 2.74s/piece,which reached the expected target.