基于焊缝兴趣点的多尺度形状描述符模板匹配算法
|
摘要
针对厢体焊接涉及多种类型焊缝,提出了一种焊缝类型识别方法.采用激光条纹提取算法获取焊缝骨架,对焊缝骨架进行直线拟合,求取焊缝骨架特征点作为兴趣点,利用多尺度形状描述符对兴趣点进行特征描述,获取焊缝类型的形状信息.选取多种类型焊缝模板,通过模板匹配算法进行焊缝类型识别.结果表明,焊缝类型分类正确率可达95%以上,这种焊缝类型识别方法是可行的.
Aiming at welding many types of welds involved in car body, an identification method of weld type was proposed. The extraction algorithm of laser fringe was used to obtain the weld seam skeleton and the weld skeleton was linearly fitted to obtain the weld skeleton feature point as a point of interest. Using multi-scale shape descriptors to characterize the interest points, the shape information of the weld types was obtained. Selecting multiple types of weld templates and identifying weld types were achieved by the template matching algorithm. The results show that the correct classification rate of weld types can be achieved up to 95%,which means this type of weld identification method was feasible.
Aiming at welding many types of welds involved in car body, an identification method of weld type was proposed. The extraction algorithm of laser fringe was used to obtain the weld seam skeleton and the weld skeleton was linearly fitted to obtain the weld skeleton feature point as a point of interest. Using multi-scale shape descriptors to characterize the interest points, the shape information of the weld types was obtained. Selecting multiple types of weld templates and identifying weld types were achieved by the template matching algorithm. The results show that the correct classification rate of weld types can be achieved up to 95%,which means this type of weld identification method was feasible.
引文
[1]何银水,孔萌,陈华斌,等.基于视觉注意机制的机器人厚板焊接焊缝轮廓的识别[J].焊接学报,2015,36(12):51-55.He Yinshui,Kong Meng,Chen Huabin,et al.Weld seam profile identification based on visual attention mechanism in robotic thick-plate welding[J].Transactions of the China Welding Institution,2015,36(12):51-55.
[2]李原,徐德,沈扬,等.一种焊缝结构光图像处理与特征提取方法[J].焊接学报,2006,27(9):25-31.Li Yuan,Xu De,Shen Yang,et al.An image processing and features extraction methods for structured light image of welding seam[J].Transactions of the China Welding Institution,2006,27(9):25-31.
[3]Zunic D,Zunic J.Shape ellipticity from Hu moment invariants[J].Applied Mathematics&Computation,2014,226:406-414.
[4]Yang J,Wang H,Yuan J,et al.Invariant multi-scale descriptor for shape representation,matching and retrieval[J].Computer Vision&Image Understanding,2016,145(C):43-58.
[5]Fan H,Yang C,Tang Y.Object detection based on scale-invariant partial shape matching[J].Machine Vision and Applications,2015,26(6):1-11.
[6]Kumar S.Object recognition using shape context with Canberra distance[J].Social Science Electronic Publishing,2017,1(2):268-273.
[7]Yang C,Feinen C,Tiebe O,et al.Shape-based object matching using interesting points and high-order graphs[J].Pattern Recognition Letters,2016,83:251-260.
[8]Ding Y,Huang W,Kovacevic R.An on-line shape-matching weld seam tracking system[J].Robotics and Computer-Integrated Manufacturing,2016,42:103-112.
[9]李原,徐德,沈扬,等.焊缝跟踪视觉传感器中图像多类型特征选择与提取[J].传感技术学报,2006,19(6):2676-2681.Li Yuan,Xu De,Shen Yang,et al.Multi type feature selection and extraction of image in seam tracking vision sensor[J].Journal of Sensing Technology,2006,19(6):2676-2681.
[10]伏喜斌,林三宝,杨春利,等.基于激光视觉传感的角焊缝外形尺寸检测[J].焊接学报,2008,29(7):47-50.Fu Xibin,Lin Sanbao,Yang Chunli,et al.Detection of the dimensions of fillet weld based on laser vision sensing[J].Transactions of the China Welding Institution,2008,29(7):47-50.
[11]Chen Q,Zhang W,Du D,et al.A new straight line detection method in images for robot seam tracking[J].China Welding,2006,15(2):1-5.
[12]Li X,Li X H,Ge S,et al.Automatic welding seam tracking and identification[J].IEEE Transactions on Industrial Electronics,2017,64(9):7261-7271.
[2]李原,徐德,沈扬,等.一种焊缝结构光图像处理与特征提取方法[J].焊接学报,2006,27(9):25-31.Li Yuan,Xu De,Shen Yang,et al.An image processing and features extraction methods for structured light image of welding seam[J].Transactions of the China Welding Institution,2006,27(9):25-31.
[3]Zunic D,Zunic J.Shape ellipticity from Hu moment invariants[J].Applied Mathematics&Computation,2014,226:406-414.
[4]Yang J,Wang H,Yuan J,et al.Invariant multi-scale descriptor for shape representation,matching and retrieval[J].Computer Vision&Image Understanding,2016,145(C):43-58.
[5]Fan H,Yang C,Tang Y.Object detection based on scale-invariant partial shape matching[J].Machine Vision and Applications,2015,26(6):1-11.
[6]Kumar S.Object recognition using shape context with Canberra distance[J].Social Science Electronic Publishing,2017,1(2):268-273.
[7]Yang C,Feinen C,Tiebe O,et al.Shape-based object matching using interesting points and high-order graphs[J].Pattern Recognition Letters,2016,83:251-260.
[8]Ding Y,Huang W,Kovacevic R.An on-line shape-matching weld seam tracking system[J].Robotics and Computer-Integrated Manufacturing,2016,42:103-112.
[9]李原,徐德,沈扬,等.焊缝跟踪视觉传感器中图像多类型特征选择与提取[J].传感技术学报,2006,19(6):2676-2681.Li Yuan,Xu De,Shen Yang,et al.Multi type feature selection and extraction of image in seam tracking vision sensor[J].Journal of Sensing Technology,2006,19(6):2676-2681.
[10]伏喜斌,林三宝,杨春利,等.基于激光视觉传感的角焊缝外形尺寸检测[J].焊接学报,2008,29(7):47-50.Fu Xibin,Lin Sanbao,Yang Chunli,et al.Detection of the dimensions of fillet weld based on laser vision sensing[J].Transactions of the China Welding Institution,2008,29(7):47-50.
[11]Chen Q,Zhang W,Du D,et al.A new straight line detection method in images for robot seam tracking[J].China Welding,2006,15(2):1-5.
[12]Li X,Li X H,Ge S,et al.Automatic welding seam tracking and identification[J].IEEE Transactions on Industrial Electronics,2017,64(9):7261-7271.