The Design Of The Tires Defect Detection System Based On X-ray

With the rapid development of the automobile industry and the increasing usage of thetires, the tire quality becomes the guarantee of the people’s property and life.So the detectionof every tire has become an important part of the tire production. The traditional software ofthe tire detection system is complex and un-useful. As the most important part, defectdetection depends on the people’s observation. This way greatly reduce the efficiency andcorrectness.This article base on the x-ray detection, which analyses the x-ray imaging principle, thecapture process of linear detector array(LDA), image acquisition and image display, imagepre-processing techniques and defect recognition. By analyzing the capture process ofU-linear detector array(U-LDA), I use multi-threading technology to realize the real-timedisplay and scrolling display for the images. In connection with the problem of large noiseand low contrast of the x-ray image, I first do the PRNU correction to stop the influence of theinherent noise, then I make some measure like median filtering, unshaped mask method, graystretch, histogram equalization method, with these measures can improve effectivesuppression of noise and the contrast. Through thorough study of the defect types and thedefect characteristic, I deny the conventional template matching method. With the failexperience, a new way, which is to design special algorithm for each kind of defect in eachregion of interest, is brought forward.To detect the abnormal array of the tires’ internal, first Ido the threshold and thinning and by acquiring the width of the tire cord to judge the imagecord sparse or not and by line scanning to judge the tire cord bend or not. To detect the blisterin the tire, which I use the maximum filtering and region growth to segment the image,finally check the result is proposed.This paper adopts the Microsoft Visual Stdio2005as the developing environment, and allthe algorithms designed have been implemented in it. A mass of X-ray tire images from thepractical production have been tested, and the experiment proves that the algorithms arefeasible and effective.