基于机器视觉的斑点图谱法机油污染度检测

邱峻伟; 王海林

doi:10.13433/j.cnki.1003-8728.2014.1017

基于机器视觉的斑点图谱法机油污染度检测

doi: 10.13433/j.cnki.1003-8728.2014.1017

1.
华南农业大学工程学院, 广州 510642

详细信息

作者简介:
邱峻伟(1987-)，硕士研究生，研究方向为车辆检测与诊断技术，qiujw@stu.scau.edu.cn;王海林(联系人)，教授，硕士生导师，hlwang@scau.edu.cn

邱峻伟(1987-)，硕士研究生，研究方向为车辆检测与诊断技术，qiujw@stu.scau.edu.cn;王海林(联系人)，教授，硕士生导师，hlwang@scau.edu.cn

计量
- 文章访问数: 183
- HTML全文浏览量: 33
- PDF下载量: 2
- 被引次数: 0
出版历程
- 收稿日期: 2013-05-10

Detection of the Engine Oil Pollution Degree Based on Machine Vision and Spot Atlas Method

1.
College of Engineering, South China Agricultural University, Guangzhou 510642

摘要

摘要: 机油滤纸斑点图谱法是判断机油污染程度的一种常用方法,但这种方法带有一定的主观性。采用机器识别的方法来代替人的判断,则可避免受到主观因素的影响。设计的机油污染度识别系统基于图像检测原理,提取了油斑图像的R、G、B、Gray、a和b这6个颜色特征值,经相关性分析得出6个特征值与机油综合污染度呈中强程度相关,且相关性极显著。建立BP神经网络分类器识别系统,该系统能够根据采集到的油斑图片判断机油的污染程度,为汽车发动机的按质换油提供技术支持。
- 机油 /
- 污染度检测 /
- 机器视觉 /
- BP神经网络
Abstract: Engine oil filter paper spot atlas method is a common method to determine the pollution degree of the engineoil, but this method has a certain degree of subjectivity.Using machine recognition method to replace human judgmentcan avoid the influence of subjective factors.The engine oil pollution degree recognition system designed in this paper isbased on digital image detection theory, in which the six colors characteristic values of R, G, B, Gray, a and b areextracted from the spots atlas image.It shows that the six colors characteristic were medium to strong correlation to thedegree of contamination and the correlation was significant.A backpropagation neural network classifier recognitionsystem was established and the system can determine the degree of contamination of engine oil according to the collectedoil spots picture, and it provides the technical support for automotive replace engine oil by qualit.
- backpropagation algorithms /
- computer vision /
- correlation methods /
- engine oil

HTML全文

参考文献(11)

[1]	樊建春，张来斌，赵雪红.在用机油斑点扩散性定量测试方法研究[J].润滑与密封,2004,161(1): 43-45Fan J C，Zhang L B，Zhao X H.Study on the visualquantitative testing of used-oil stain dispersiveness[J].Lubrication Engineering，2004，161 (1): 43-45 (in Chinese)
[2]	吉国光.汽车发动机在用机油斑点图谱检测法[J].汽车维修技师,2010，(3): 83Ji G G.Detection method of vehicle used-oil spot [J].Auto Maintenance,2010，(3): 83 (in Chinese)
[3]	邓继忠，张泰岭.数字图像处理技术[M].广州: 广东科技出版社,2005Deng J Z,Zhang T L.Digital image processingtechniques[M].Guangzhou: Guangdong Science &Technology Press,2005 (in Chinese)
[4]	王焱，彭慧玲，马勇.一种彩色图像去噪的方法[J].微计算机信息,2010,26(9 /2): 23-24Wang Y，Peng H L，Ma Y.A method of color imagedenoising [J].Microcomputer Information,2010，26(9 /2): 23-24 (in Chinese)
[5]	Rafael C.Gonzalez，Richard E，Woods，et al.Digitalimage processing,second edition [M].Beijing:Publishing House of Electronics Industry,2003
[6]	宋寅卯，李迎成.彩色图像分割技术进展与展望[J].电脑开发与应用,2012,25(7): 26-28Song M Y，Li Y C.Progress and prospects of the colorimage segmentation technology [J].ComputerDevelopment & Applications，2012，25 (7): 26-28 (in Chinesee)
[7]	Carevic D，Caelli T.Region-based coding of color imageusing karhunen-loeve transform [J].Graphics Modelsand Image Processing,1997,59(1): 27-38
[8]	Bezdek J C.Pattern recognition with fuzzy objectivefunction algorithms [M].New York: PlenumPress，1981
[9]	王瑞.应用机油斑点图谱法进行按质换油的探讨[J].设备管理与维修,1993,107(9): 27-28Wang R.Disscussion on engine oil change according tothe quality by using oil spot method [J].PlantMaintenance Engineering，1993，107 (9): 27-28 (in Chinesee)
[10]	邱峻伟.基于机器视觉的汽车发动机润滑油污染度检测[D].广州: 华南农业大学,2013Qiu J W.Detection of automotive engine lubricant oilpollution degree based on the machine vision[D] .Guangzhou: South China Agricultural University，2013(in Chinese)
[11]	刘天舒.BP 神经网络的改进研究及应用[D].哈尔滨: 东北农业大学，2011Liu T S.The Research and application on BP neuralnetwork improvement [D].Harbin: NortheastAgricultural University,2011 (in Chinese)