Tire Belt Layer Defect Detection Based On X-ray Imaging And Texture Analysis

At present,with the development of economy,the use of automobiles has increased dramatically,so the safety of automobiles has attracted more and more attention.As an important part of automobiles,the quality of tires is also very important.So tires have to be inspected strictly before they leave the factory,but at present most of the major domestic tire factories adopt foreign tire inspection equipment.And the import price of these equipment is very expensive.In addition,the criteria for defect grading at home and abroad are also different.Therefore,there is an urgent need to develop an automatic tire defect detection system in China.Based on this situation,an automatic tire defect detection system is designed and developed.This paper focuses on the design and implementation of the defect detection algorithm in the band layer,and finally optimizes the algorithms and integrate into the system.The main research contents are as follows:(1)The structure and working principle of the existing X-ray tire detection system are systematically studied,and the X-ray imaging principle of tires is analyzed.(2)Combining with the characteristics of the texture of the band layer,the commonly texture defect detection algorithms are analyzed.(3)An algorithm for segmentation of tire band layer is designed.Algorithms are designed according to the texture characteristics for six types of defects in the band layer.Design the detection algorithm of losing the 0 degree band defect.By using morphological filtering,the kernel function of morphological is defined to retain the cord of 0 degree,while the other direction cords are filtered to detect the existence of 0 degree band layer.The second and zero band layer offset defect detection algorithm is designed.The boundary of the second band layer can be obtained when dividing the band layer,and the zero band layer boundary can be obtained when detecting the defect of the zero band layer.The existence of the second band layer and the zero band layer offset can be judged by comparing the difference between the two boundaries.The detection algorithm of overlapping and cracking defects of band layer is designed.Using the idea of rotating projection,the projection images of all rotating angles are traversed,and the abnormal peaks or valleys in the projection images are detected to determine whether there are overlaps or cracks in the band layer.An impurity defect detection algorithm for band layer is designed.Firstly,a local binarization algorithm is defined,and then an area filter is made on the binarized image,so that the larger area of impurities is retained and other textures are filtered.Design an algorithm for band layer discount.The histogram information of the band layer in the small window is calculated,and the discount of the band layer is judged by comparing the similarity of the histogram in the adjacent window.(4)The algorithm is integrated into the tire automatic defect detection system.Defect detection algorithm applies Visual Studio 2015 software development environment.The development of automatic tire defect detection system is based on Windows platform.