基于非负矩阵分解算法的工程陶瓷磨削表面损伤检测

林滨; 张彦斌; 陈善功

doi:10.3788/OPE.20122011.2459

您当前的位置：

首页 >

文章列表页 >

基于非负矩阵分解算法的工程陶瓷磨削表面损伤检测

微纳技术与精密机械 | 更新时间：2020-08-12

- 基于非负矩阵分解算法的工程陶瓷磨削表面损伤检测
- Damage detection of engineering ceramics ground surface based on NMF
- 光学精密工程 2012年20卷第11期页码：2459-2464
- 作者机构：
  
  天津大学先进陶瓷与加工技术教育部重点实验室天津,300072
- 作者简介：
- 基金信息：
  
  国家自然科学基金资助项目(No. 51075296)();国家科技重大专项资助项目(No.2011ZX04002-161)()
- DOI：10.3788/OPE.20122011.2459
  中图分类号： TB30;TQ174.1+2
- 收稿日期：2012-06-03，
  
  修回日期：2012-08-30，
  
  纸质出版日期：2012-11-10
- 稿件说明：
移动端阅览
林滨, 张彦斌, 陈善功. 基于非负矩阵分解算法的工程陶瓷磨削表面损伤检测[J]. 光学精密工程, 2012,20(11): 2459-2464

LIN Bin, ZHANG Yan-bin, CHEN Shan-gong. Damage detection of engineering ceramics ground surface based on NMF[J]. Editorial Office of Optics and Precision Engineering, 2012,20(11): 2459-2464
林滨, 张彦斌, 陈善功. 基于非负矩阵分解算法的工程陶瓷磨削表面损伤检测[J]. 光学精密工程, 2012,20(11): 2459-2464 DOI： 10.3788/OPE.20122011.2459.

LIN Bin, ZHANG Yan-bin, CHEN Shan-gong. Damage detection of engineering ceramics ground surface based on NMF[J]. Editorial Office of Optics and Precision Engineering, 2012,20(11): 2459-2464 DOI： 10.3788/OPE.20122011.2459.

摘要

考虑表面损伤检测在工程陶瓷表面质量评价中的重要作用

首次把非负矩阵分解(NMF)图像重构算法引入工程陶瓷磨削表面损伤检测中

并进行了理论分析与实例检测。首先

将输入图像数据集从原始数据空间降维到一个低维NMF空间

利用本文提出的图像重构相对误差0.1监督规则

确定子空间基

值。然后

利用两个低维非负矩阵进行图像重构

获取磨削纹理背景图像

并通过图像减法去除磨削纹理。最后

利用Canny边缘检测算法提取工程陶瓷磨削表面损伤图像。实验结果表明

该方法能够准确提取表面损伤并计算磨削损伤率评价参数。

Abstract

As the surface damage detection plays an important role in evaluating engineering ceramic surface quality

this paper introduces an image reconstruction algorithm

Nonnegative Matrix Factorization (NMF) algorithm

into the damage detection of engineering ceramics grounding surface for the first time. It analyzes the theoretical function of the algorithm and gives a detection example. First

the input image data set was reduced from an original data space to a lower-dimensional NMF space

and the image reconstruction relative error 0.1 rule proposed by this paper was used to determine a proper space basis

value. Then

the background image of ground texture was obtained by image reconstruction using two lower-dimensional nonnegative matrixes

and the ground textures were removed by image subtraction. Finally

the Canny edge detection was used to extract the damage image of engineering ceramics grounding surface. Experimental results indicate that the proposed method can accurately extract the surface damage of engineering ceramics and can calculate the evaluation parameter of grinding damage rate.

关键词

Keywords

references

ZHANG B,ZHENG X L,TOKURA H,et al.. Grinding induced damage in ceramics [J]. Journal of Materials Processing Technology, 2003, 132: 353-364.[2] KUMAR A. Computer-vision-based fabric defect detection: a survey [J]. IEEE Trans Ind Electron, 2008, 55(1): 348-363.[3] XIE X. A review of recent advances in surface defect detection using texture analysis techniques [J]. Electron Lett Comput Vis Image Anal, 2008, 7(3): 1-22.[4] GUILLAMET D,VITRI J. Non-negative matrix factorization for face recognition [J]. Lect Notes Comput Sci, 2002, 2504: 336-344.[5] LEE D D,SEUNG H S. Learning the parts of objects by non-negative matrix factorization [J]. Nature, 1999, 401(10): 788-791.[6] GUILLAMET D,VITRIA J,SCHIELE B. Introducing a weighted non-negative matrix factorization for image classification [J]. Int Conf Pattern Recognit, 2002, 2: 116-119.[7] GUILLAMET D,VITRI J,SCHIELE B. Introducing a weighted non-negative matrix factorization for image classification [J]. Pattern Recognition Letters, 2003, 24: 2447-2454.[8] 高涛,何明一. 改进投影梯度非负矩阵分解的单训练样本特征提取研究[J]. 电子与信息学报, 2010, 32(5): 1121-1125. GAO T,HE M Y. Using improved non-negative matrix factorization with projected gradient for single-trial feature extraction [J]. Journal of Electronics & Information Technology, 2010, 32(5): 1121-1125. (in Chinese)[9] 王梁,郝燕玲,张振兴. 基于NMF闭塞字典的压缩传感声纳图像识别[J]. 华中科技大学学报:自然科学版, 2011, 39(9): 29-33. WANG L,HAO Y L,ZHANG ZH X. Sonar image recognition of compressed sensing using NMF occlusion dictionary [J]. J. Huazhong Univ. of Sci. & Tech.:Natural Science Edition, 2011, 39(9): 29-33. (in Chinese)[10] LEE D D,SEUNG H S. Algorithms for non-negative matrix factorization [J]. Adv Neural Inf Process Syst, 2001, 13: 556-562.[11] 李乐,章毓晋. 基于线性投影结构的非负矩阵分解[J]. 自动化学报, 2010, 36(1): 23-39. LI L, ZHANG Y J. Linear Projection-based Non-negative Matrix Factorization [J]. Acta Automatica Sinica, 2010, 36(1): 23-39. (in Chinese)[12] GONG P H, ZHANG C S. Efficient nonnegative matrix factorization via projected newton method [J]. Pattern Recognition, doi:10.1016/j.patcog.2012.02.037.[13] LIN C J. Projected gradient methods for non-negative matrix factorization [J]. Neural Comput, 2007, 19(10): 2756-277.[14] 张俊雄,苟一,李伟. 基于形态特征的玉米种子表面裂纹检测方法[J]. 光学精密工程, 2007,15(6): 951-956. ZHANG J X,GOU Y,LI W. Detection of surface cracks of corn kernel based on morphology [J]. Opt. Precision Eng., 2007, 15(6): 951-956. (in Chinese)

浏览量

754

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

切向载荷作用下氮化硅陶瓷崩碎损伤规律与机理