基于穿线法的胎侧散线缺陷检测
|
摘要
针对轮胎内部胎侧出现的散线缺陷,提出一种基于穿线法的胎侧散线缺陷检测方法;该方法利用大津法和快速并行细化算法对X射线轮胎图像分别进行二值化、细化等预处理,以利于观察轮胎帘线信息;根据细化后胎侧的帘线确定上、下2条边界线,并在2条边界线内以一定步长进行穿线,进而统计穿线与帘线之间交点总个数以及每条穿线上交点的平均个数、相邻交点疏密情况;通过每条穿线上的交点个数与之前所求平均个数进行比较,结合相邻交点的疏密情况、有无交叉点,进而判断是否存在散线缺陷;对1 000幅包含有人为添加不同弯曲程度、斜率走向的散线细化图片和168张实际的X射线轮胎图片分别进行缺陷检测。结果表明,该方法可以有效地检测出胎侧部位的散线缺陷并标记其所在区域。
Aiming at the defect of existing in loose line the sidewall of tires,a method for defect detection of sidewall loose line based on threading method was proposed. X ray tire images were binarized and refined by the Otsu method and the fast parallel refinement algorithm separately,which was beneficial to observation of the tire cord information. According to the cord of the refinement sidewall,the upper and lower boundary lines were determined,and threading was operated in the two boundary lines with a certain length,the purpose for the arrangement was to calculate the total number of intersections between the threading and cord,the average number of the intersection and the density of the adjacent intersection on each thread. In order to judge the existence of loose line defects,the number of intersections on each threading line was compared with the previously calculated average number,and the density and crossing condition of adjacent points of intersection were considered. 1 000 binarized images artificially added varying degrees of bending and different slope of loose line and 168 actual tire images were tested respectively. The experimental results show that the proposed method can effectively detect the defect area of the tire and draw the area of loose line.
Aiming at the defect of existing in loose line the sidewall of tires,a method for defect detection of sidewall loose line based on threading method was proposed. X ray tire images were binarized and refined by the Otsu method and the fast parallel refinement algorithm separately,which was beneficial to observation of the tire cord information. According to the cord of the refinement sidewall,the upper and lower boundary lines were determined,and threading was operated in the two boundary lines with a certain length,the purpose for the arrangement was to calculate the total number of intersections between the threading and cord,the average number of the intersection and the density of the adjacent intersection on each thread. In order to judge the existence of loose line defects,the number of intersections on each threading line was compared with the previously calculated average number,and the density and crossing condition of adjacent points of intersection were considered. 1 000 binarized images artificially added varying degrees of bending and different slope of loose line and 168 actual tire images were tested respectively. The experimental results show that the proposed method can effectively detect the defect area of the tire and draw the area of loose line.
引文
[1]郭奇.基于X射线轮胎缺陷检测系统设计[D].太原:中北大学,2015.
[2]HUANG Z H,LIU Z,ZHU M,et al.Defects on-line detection of tire textures based on statistical features[J].Optical Technique,2009,35(1):32-60.
[3]XIANG Y,ZHANGC,GUO Q.A dictionary-based method for tire defect detection[C]//IEEE International Conference on Information and Automation,26-31 July 2014,Hailar,China.Piscataway:IEEE RAS,2014:519-523.
[4]XIANG Y Y.Tire defect detection using local and global features[J].Computing Aided Drafting,Design and Manufacturing,2013,23(4):49-52.
[5]冯霞,石超,丁文波,等.基于傅里叶变换的频谱分析法在X射线轮胎检测中的应用[J].CT理论与应用研究,2014,23(3):453-458.
[6]宋开勇,王传旭,董向鹏.轮胎纹理提取与缺陷检测方法[J].计算机工程,2012,38(18):291-293.
[7]赵成忠,龙小丽.全钢载重子午线轮胎0°带束层散线原因分析及解决措施[J].轮胎工业,2009,29(3):176-178.
[8]高鹏.基于X光图像的轮胎内部缺陷检测技术研究[D].天津:天津大学,2009.
[9]魏畅然.基于穿线法的数字识别方法[J].科技情报开发与经济,2010,20(28):137-139.
[10]杨陶,田怀文,刘晓敏,等.基于边缘检测与Otsu的图像分割算法研究[J].计算机工程,2016,42(11):255-260.
[11]吕岳,施鹏飞.一种实用并行细化算法及其实现[J].计算机工程与设计,2000,21(4):53-56.
[12]牟少敏,杜海洋,苏平,等.一种改进的快速并行细化算法[J].微电子学与计算机,2013,30(1):53-55.
[2]HUANG Z H,LIU Z,ZHU M,et al.Defects on-line detection of tire textures based on statistical features[J].Optical Technique,2009,35(1):32-60.
[3]XIANG Y,ZHANGC,GUO Q.A dictionary-based method for tire defect detection[C]//IEEE International Conference on Information and Automation,26-31 July 2014,Hailar,China.Piscataway:IEEE RAS,2014:519-523.
[4]XIANG Y Y.Tire defect detection using local and global features[J].Computing Aided Drafting,Design and Manufacturing,2013,23(4):49-52.
[5]冯霞,石超,丁文波,等.基于傅里叶变换的频谱分析法在X射线轮胎检测中的应用[J].CT理论与应用研究,2014,23(3):453-458.
[6]宋开勇,王传旭,董向鹏.轮胎纹理提取与缺陷检测方法[J].计算机工程,2012,38(18):291-293.
[7]赵成忠,龙小丽.全钢载重子午线轮胎0°带束层散线原因分析及解决措施[J].轮胎工业,2009,29(3):176-178.
[8]高鹏.基于X光图像的轮胎内部缺陷检测技术研究[D].天津:天津大学,2009.
[9]魏畅然.基于穿线法的数字识别方法[J].科技情报开发与经济,2010,20(28):137-139.
[10]杨陶,田怀文,刘晓敏,等.基于边缘检测与Otsu的图像分割算法研究[J].计算机工程,2016,42(11):255-260.
[11]吕岳,施鹏飞.一种实用并行细化算法及其实现[J].计算机工程与设计,2000,21(4):53-56.
[12]牟少敏,杜海洋,苏平,等.一种改进的快速并行细化算法[J].微电子学与计算机,2013,30(1):53-55.