基于特征直方图的轮胎X光图像杂质检测算法

Impurity Detection Algorithm of Tire X-Ray Image Based on Feature Histogram

为提高轮胎生产中对杂质材料的检测能力,消除轮胎出厂使用时的安全隐患,提出一种基于图像灰度分布曲线的杂质缺陷检测算法。算法先通过比较轮胎X光图像不同区域杂质位置与正常位置灰度分布曲线的不同,沿垂直方向上选取反映杂质缺陷的特征值并统计其直方图,再根据直方图采用最大间距法设置自适应阈值进行缺陷检测,最终根据不同材质的杂质缺陷在灰度分布曲线上的不同特征将杂质分为金属杂质和非金属杂质。算法经实验验证能够准确检测出杂质缺陷,在准确性上相比已有方法有所提高,具有一定的实际应用价值。

In order to improve the detection ability of impurity materials in tire production and elimi-nate the potential safety hazard when tires are used in factory,an impurity defect detection algorithm based on image gray distribution curve is proposed.Firstly,the algorithm compares the gray distribution curves of impurity positions in different regions of tire X-ray images with those in normal positions,selects the characteristic values reflecting impurity defects along the vertical direction and counts their histograms,then sets an adaptive threshold value according to the histogram with the maximum spacing method to detect defects,and finally divides impurities into metal impurities and non-metal impurities according to the different characteristics of impurity defects in different materials on the gray distribution curves.The experimental results show that the algorithm can accurately detect impurity defects,and the accuracy is improved compared with the existing methods,which has certain practical application value.