欢迎访问润滑与密封官方网站!

咨询热线:020-32385313 32385312 RSS EMAIL-ALERT
基于显微图像识别的在线滑油中磨粒分析方法
作者:
作者单位:

作者简介:

通讯作者:

中图分类号:

基金项目:

国家自然科学基金项目(E091004)


Online Analysis for Particles in Lubricating Oil Based on Microimage Recognition Method
Author:
Affiliation:

Fund Project:

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    针对现有显微图像油液磨粒监测技术的芯片结构的不足,改进微通道与外通道结合处结构,减小了微通道被阻塞的可能性,提高了检验精度。为了提高机器润滑油中金属磨粒在线图像识别的准确性,分析图像运动模糊退化模型,并以自相关函数估算模糊尺度,采用基于维纳滤波的图像恢复算法获得清晰的磨粒图像;应用图像差值与最大类间方差法(Otsu)对图像进行分割并对分割后的磨粒图像计数获得磨粒等效直径和分布。研究结果表明,该图像分割方法提高了油液磨粒的在线实时检测的准确率,其精度达95%以上。

    Abstract:

    Aimed at the shortage of microfluidic chip structure,the channel junction’s structure was improved,the likelihood was decreased that the micro channel was blocked,and the inspection accuracy was improved.To improve the accuracy of the lubricating oil pollution level analysis and the metal particle image identification,all motion blur image degradation model was analyzed; the autocorrelation function was used to estimate the fuzzy measurement;the second Wiener filter was applied to restore blurred image;the wear particle image segmentation method based on lubricating oil background image and Otsu was applied;the last the numbers of the particles were counted and the equivalent diameter was calculated. Experiments show that the online realtime detection system for oil detection has very good effect. Its precision is above 95%.

    参考文献
    相似文献
    引证文献
引用本文

郝延龙,何红坤,常青,严志军,潘新祥,朱新河,程东,马春生.基于显微图像识别的在线滑油中磨粒分析方法[J].润滑与密封,2016,41(5):59-64.
HAO Yanlong, HE Hongkun, CHANG Qing, YAN Zhijun, PAN Xinxiang, ZHU Xinhe, CHENG Dong, MA Chunsheng. Online Analysis for Particles in Lubricating Oil Based on Microimage Recognition Method[J]. Lubrication Engineering,2016,41(5):59-64.

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:
  • 最后修改日期:
  • 录用日期:
  • 在线发布日期: 2016-06-16
  • 出版日期: