基于YOLOX模型的轮胎X射线图像中0°带束层接头检测  

0°Belt Joint Detection in Tire X-ray Images Based on YOLOX Model

作　　者：史建杰 李金屏[2] 赵建玉[1] SHI Jianjie;LI Jinping;ZHAO Jianyu(School of Electrical Engineering,University of Jinan,Jinan 250022,Shandong,China;School of Information Science and Engineering,University of Jinan,Jinan 250022,Shandong,China)

机构地区：[1]济南大学自动化与电气工程学院,山东济南250022 [2]济南大学信息科学与工程学院,山东济南250022

出　　处：《济南大学学报（自然科学版）》2023年第3期280-287,共8页Journal of University of Jinan(Science and Technology)

基　　金：山东省重点研发计划项目(2017CXGC0810)。

摘　　要：针对现有轮胎X射线图像中0°带束层接头检测算法易受轮胎自身花纹影响,造成较严重的误报和漏报的问题,提出基于YOLOX模型的轮胎X射线图像中0°带束层接头检测方法;利用投影直方图定位带束层区域,缩小检测范围,提高检测速度;在自建数据集上测试所提方法的检测性能。结果表明,所提出的方法能够准确定位0°带束层接头,误报率、漏报率分别小于1.15%、0.59%。Aiming at the problem that 0°belt joint detection algorithms in existing tire X-ray images were easily affected by patterns of tires,which leaded to serious false positives and missing positives,a method of 0°belt joint detection based on YOLOX model was proposed.Projection histograms were used to locate the belt layer region,narrow the detection range,and improve the detection speed.Detection performances of the proposed method were tested on a self-built data set.The results show that the proposed method can accurately locate 0°belt joints,and the false positive rate and missing negative rate are less than 1.15%and 0.59%,respectively.

关 键 词：轮胎 0°带束层接头缺陷 目标检测 区域分割 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]