Research On Inspecting Algorithm For Stomata's Surface Defects Of Inner Tube Based On Machine Vision

Vision inspection has become an essential link in modern industrial manufacturing.As a useful detection method,machine vision contains image processing and analysis,artificial-intelligence,mode recognition.They are widely used in the applications of detection of industry products,particularly the detection of surface defects.In the industrial production of inner tube,due to uneven heating stomata manufacturing device,inadequate temperature and abnormal dusting device,it is easily to lead the inner tube pores not being pierced,stomata edge cracking,large number of powder and other defects which would seriously affect the quality of the product.The aim of this thesis is to fulfill the urgent requirement of inspecting surface defects on stomata of tire inner tube.The main works was as follows:(1)Based on the working principle of machine vision equipment for striped steel,leather and thin film's surface detection,combined with tire inner tube production process and workshop environment,we construct the software and hardware structure of the reflective inner tube hole surface detection system.(2)In this paper,an image brightness automatic adjustment algorithm,which based on the image pixel,combining linear variation and gamma change is proposed.The effect of this algorithm has been confirmed.(3)Combining the traditional threshold method and the spatial analysis,a binary segmentation method is designed to fit the surface defects of the car tire.It is proved to separate the target and the background more efficiently.Moreover,mathematical morphology filter is utilized in order to wipe off the isolated noisy spots during image segmentation.(4)Feature parameters(perimeter,area,grayscale and the dispersion of the contour points)can be extracted.The defects would be classified,shown with the analysis result on user's interface.(5)Designed the system software and interface according to the need of detection function by manufacture factory.Through the testing of a large number of different batches of inner tubes,the feasibility and validity of the algorithm and detection system have been verified correctly.