基于整体与局部纹理特征加权融合的港机装备钢丝绳断丝缺陷检测研究  

Research on broken wire defect detection of wire rope for port machinery equipment based on weighted fusion of global and local texture features

作　　者：张卫国[1] 刘聪 曾祥堃[2] 夏立成 王紫阳 ZHANG Weiguo;LIU Cong;ZENG Xiangkun;XIA Licheng;WANG Ziyang(Logistics Engineering College,Shanghai Maritime University,Shanghai 201306,China;College of Information Engineering,Shanghai Maritime University,Shanghai 201306,China)

机构地区：[1]上海海事大学物流工程学院,上海201306 [2]上海海事大学信息工程学院,上海201306

出　　处：《中国工程机械学报》2024年第3期398-403,共6页Chinese Journal of Construction Machinery

摘　　要：钢丝绳是港机装备的重要部件,由于作业工况环境恶劣,钢丝绳表面极易引起断丝等缺陷,影响港机装备作业安全。针对港机装备钢丝绳表面油泥严重、光照亮度不均,以及港机装备钢丝绳断丝缺陷能够在钢丝绳股顶钢丝椭圆形区域内有效体现的特点,提出了一种基于钢丝绳整体纹理特征与股顶钢丝椭圆形区域边缘轮廓纹理特征加权融合的钢丝绳断丝缺陷检测方法。首先采用图像降噪、增强、校正技术对钢丝绳原始图像进行预处理。然后采用图像平滑、阈值分割及边缘特征提取技术对股顶钢丝椭圆形区域边缘轮廓进行提取。接着采用局部二值模式(LBP)算子分别提取钢丝绳整体纹理特征与股顶钢丝椭圆形区域边缘轮廓纹理特征,并对钢丝绳整体纹理及边缘轮廓纹理特征进行特征加权融合。最后对加权融合后的特征向量进行主成分分析(PCA)法降维,并应用支持向量机(SVM)技术对钢丝绳断丝缺陷进行检测。研究结果表明:本文提出的方法对实际工况下重油泥、光照不均等钢丝绳断丝缺陷具有较好的检测效果,具有一定的工程应用价值。Wire rope is an important part of port machinery equipment.Due to the harsh working conditions,the surface of the wire rope is prone to defects such as broken wires,which affects the operation of port machinery equipment.In view of the serious sludge on the surface of the wire rope of port machinery equipment,uneven illumination brightness,and the fact that the broken wire defect of the wire rope of port machinery equipment can be effectively reflected in the elliptical areas of the wire rope strand top,this paper proposes a wire rope broken wire defect detection method based on the weighted fusion of the overall texture feature of the wire rope and the edge contour texture feature of the elliptical area of the wire rope strand top.Firstly,image noise reduction,image enhancement,and image correction technologies are used to preprocess the original image of the wire rope;then image smoothing,threshold segmentation,and edge feature extraction technologies are used to extract the edge contour of the wire rope elliptical area on the top of the strand;then the local binary patterm(LBP) operator is adopted to extract the overall texture features of the wire rope and the edge contour texture features of the elliptical area of the wire rope strand top respectively,and then the overall texture and edge contour texture features of the wire rope are organically weighted and fused;finally,principal components analysis(PCA) is used to reduce the dimensionality of the weighted and fused eigenvectors,and support vector machine(SVM) technology is adopted to detect the broken wire defect of the wire rope.The research results show that the method proposed in this paper has a better detection effect on the images of broken wire defects such as heavy oil sludge and uneven illumination under actual working conditions,and has good engineering application value.

关 键 词：港机装备 钢丝绳断丝检测 股顶钢丝 椭圆形轮廓 纹理特征 特征加权融合 

分 类 号：U653.92[交通运输工程—港口、海岸及近海工程]