基于机器视觉的汽车线束外观检测方法研究
|
摘要
针对汽车线束端子及线束线材在生产过程中的压接尺寸、线序错误等典型缺陷进行研究,对于端子压接的线材尺寸缺陷,利用Blob分析识别定位检测区域并分割出铜线区域,再提取区域几何特征检测该缺陷;对于线束漏接错接等缺陷,通过NCC(归一化互相关)模板匹配识别定位须检测的区域并分割出不同颜色线束的区域,提取颜色矩特征,建立基于MLP(多层感知机)的颜色识别模型对线束颜色识别分类并检测缺陷,最后通过C#和HALCON的联合编程实现。
Studies the typical defects such as the crimping quality of the automobile wire harness terminal and the mismatch of wire harness connector in the process of manufacturing. For the defect of the single wire terminal, the Blob analysis is used to identify the region of interest area and divide the copper wire area, and then the geometrical feature is used to detect the defect. For the defects such as the mismatching of the wire harness, the NCC(Normalized Cross-Correlation) template matching is used to identify and locate the region to be detected and segment the regions of different color harnesses, then extracts the color moment features and establish the color recognition model by multilayer perceptron classifies, and finally implement the application by C# and HALCON.
Studies the typical defects such as the crimping quality of the automobile wire harness terminal and the mismatch of wire harness connector in the process of manufacturing. For the defect of the single wire terminal, the Blob analysis is used to identify the region of interest area and divide the copper wire area, and then the geometrical feature is used to detect the defect. For the defects such as the mismatching of the wire harness, the NCC(Normalized Cross-Correlation) template matching is used to identify and locate the region to be detected and segment the regions of different color harnesses, then extracts the color moment features and establish the color recognition model by multilayer perceptron classifies, and finally implement the application by C# and HALCON.
引文
[1]薛建凌,蔡振军.我国线束线缆行业的可持续发展道路[J].管理观察,2017(9):48-49.
[2]闫亮亮,谷孝卫,姬红亮,等.汽车线束保护介绍与分析[J].汽车电器,2013(8):19-22.
[3]胡凯,李臻,李延静.汽车线束布置的研究[J].汽车零部件,2017(12).50-52.
[4]邹松明.端子压接工艺与服务过程中的问题分析与改进研究[D].东华大学,2014.
[5]海潮,赵凤霞,孙烁.基于Blob分析的红枣表面缺陷在线检测技术[J].食品与机械,2018,34(01):126-129.
[6]任楷飞,孟令军,顾泽凌.基于灰度值金字塔算法的零件识别系统设计[J].中国测试,2018,44(07):83-87.
[7]曾文锋,李树山,王江安.基于仿射变换模型的图像配准中的平移、旋转和缩放[J].红外与激光工程,2001(01):18-20+17.
[8]张少博,全书海,石英,等.基于颜色矩的图像检索算法研究[J].计算机工程,2014,40(6):252-255.
[9]Hecht-Nielsen R.Theory of the Backpropagation Neural Network[J]. Neural Networks,1988,1(1):445-445.
[2]闫亮亮,谷孝卫,姬红亮,等.汽车线束保护介绍与分析[J].汽车电器,2013(8):19-22.
[3]胡凯,李臻,李延静.汽车线束布置的研究[J].汽车零部件,2017(12).50-52.
[4]邹松明.端子压接工艺与服务过程中的问题分析与改进研究[D].东华大学,2014.
[5]海潮,赵凤霞,孙烁.基于Blob分析的红枣表面缺陷在线检测技术[J].食品与机械,2018,34(01):126-129.
[6]任楷飞,孟令军,顾泽凌.基于灰度值金字塔算法的零件识别系统设计[J].中国测试,2018,44(07):83-87.
[7]曾文锋,李树山,王江安.基于仿射变换模型的图像配准中的平移、旋转和缩放[J].红外与激光工程,2001(01):18-20+17.
[8]张少博,全书海,石英,等.基于颜色矩的图像检索算法研究[J].计算机工程,2014,40(6):252-255.
[9]Hecht-Nielsen R.Theory of the Backpropagation Neural Network[J]. Neural Networks,1988,1(1):445-445.