| 基于图像分析的滚动轴承表面缺陷识别技术研究 |
| |
| 引用本文： | 杨阳,蒋先刚.基于图像分析的滚动轴承表面缺陷识别技术研究[J].华东交通大学学报,2008,25(6):41-46. |
| |
| 作者姓名： | 杨阳 蒋先刚 |
| |
| 作者单位： | 1. 华东交通大学,机电工程学院,江西,南昌,330013
2. 华东交通大学,基础科学学院,江西,南昌,330013 |
| |
| 摘 要： | 系统介绍了滚动轴承表面缺陷视觉检测系统的检测原理及流程，着重分析了表面缺陷二值图像的几何特征和形状矩特征、BP神经网络的各参数选择以及实验结果．实验证明本系统用于滚动轴承表面缺陷检测具有可行性，采用神经网络识别可使表面缺陷分类正确率达到88％．
|
| 关 键 词： | 几何特征 形状矩特征 BP神经网络 |
The Research on Recognition of Rolling Bearing Surface Defects Based on Image Analysis |
| |
| YANG Yang,JIANG Xian-gang.The Research on Recognition of Rolling Bearing Surface Defects Based on Image Analysis[J].Journal of East China Jiaotong University,2008,25(6):41-46. |
| |
| Authors: | YANG Yang JIANG Xian-gang |
| |
| Institution: | YANG Yang1,JIANG Xian-gang2(1.School of Mechanical , Electrical Engineering,2.School of Basic Science,East China Jiaotong Universiey,Nancharg 330013,China) |
| |
| Abstract: | The paper introduces the principle and flow concerning the automatic detection system of rolling bearing surface defects,and analyzes that the general geometry features,moment features and the parameter choices in the BP neural network.Experiment shows its feasibility and the accuracy rate of the surface defect categories amounts to 88 percent. |
| |
| Keywords: | geometry features moment features BP neural network |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
| 点击此处可从《华东交通大学学报》浏览原始摘要信息 |
| 点击此处可从《华东交通大学学报》下载免费的PDF全文 |
