HUANG Yonghua,LIN Zhenheng,CHEN Xuejun.A surface crack detection method for frictionblock of brake pad based on SVM[J].Journal of Yanbian University,2019,45(02):175-180.
一种基于SVM的刹车蹄块片摩擦块表面裂纹检测法
- Title:
- A surface crack detection method for friction block of brake pad based on SVM
- 文章编号:
- 1004-4353(2019)02-0175-06
- Keywords:
- crack detection; brake pads; friction block; support vector machine
- 分类号:
- TP391.4
- 文献标志码:
- A
- 摘要:
- 针对刹车片外观裂纹检测需求,通过构建刹车蹄块片图像采集系统,提出了一种基于支持向量机SVM的刹车蹄块片摩擦块表面裂纹检测法.该方法首先利用灰度图像的梯度模值信息,投影提取摩擦块所在区域; 然后以局部窗口子图像为单位,计算灰度共生矩阵并提取相关特征量; 最后采用分类样本对支持向量机分类模型进行训练,对摩擦块表面裂纹缺陷和正常区域进行分类预测.实验表明:该方法能较好地实现摩擦块表面裂纹缺陷和正常区域的分类,对表面裂纹缺陷存在与否的判定准确率可达98.33%.
- Abstract:
- Aiming at the need of surface crack detection of brake pad, an image acquisition system of brake pad is constructed, and a surface crack detection method of brake friction pad is proposed based on support vector machine(SVM). Firstly, the gradient modulus information of the gray image is used to project and extract the friction block region. Then, the gray level co-occurrence matrix is calculated and the relevant feature quantities are extracted based on the local window sub-images. Finally, the support vector machine classification model is trained with the classified samples to classify and predict the surface crack defects and normal regions of the friction block. Experiments show that the method can better classify the surface crack defects and normal areas of friction blocks, and the accuracy of judging whether surface crack defect exists or not can reach 98.33%.
参考文献/References:
[1] 徐冬,杨荃,王晓晨,等.基于机器视觉的热轧中间坯镰刀弯在线检测系统[J].中南大学学报(自然科学版),2018,49(7):1657-1666.
[2] BAI Xuefei, WANG Wenjian. Principal pixel analysis and SVM for automatic image segmentation[J]. Neural Computing & Applications, 2016,27(1):45-58.
[3] 刘磊,王冲,赵树旺,等.基于机器视觉的太阳能电池片缺陷检测技术的研究[J].电子测量与仪器学报,2018,32(10):47-52.
[4] CHEN Dian, PAN Ming, HUANG Wei, et al. The provenance of nephrite in China based on multi-spectral imaging technology and gray-level co-occurrence matrix[J]. Analytical Method, 2018,10(33):4058-4059.
[5] ANCY C A, NAIR L S. Tumour classification in graph-cut segmented mammograms using GLCM features-fed SVM[J]. Intelligent Engineering Informatics, 2018,695:197-208.
[6] XIAN Guangming. An identification method of malignant and benign liver tumors from ultrasonography based on GLCM texture features and fuzzy SVM[J]. Expert Systems with Applications, 2010,37(10):6737-6741.
[7] XIONG Wei, XU Jingjing, XIONG Zijie, et al. Degraded historical document image binarization using local features and support vector machine(SVM)[J]. Optik, 2018,164:218-223.
[8] HARIKUMAR R, KARTHICK G, VINOTH K B. Earlier detection of cancer regions from MR image features and SVM classifiers[J]. International Journal of Imaging Systems and Technology, 2016,26(3):196-208.
[9] CHANG Chihchung, LIN Chihjen. LIBSVM: a library for support vector machines[J]. ACM Transactions on Intelligent Systems and Technology, 2011,2(3):1-39.
[10] 王小川,史峰,郁磊,等.MATLAB神经网络30个案例分析[M].北京:北京航空航天大学出版社,2013:122-129.
[11] 张建军,罗静.基于改进Sobel算子的表面裂纹边缘检测算法[J].合肥工业大学学报(自然科学版),2011,34(6):845-847.
[12] 闫会朋,杨正伟,田干,等.铁磁材料表面裂纹的涡流热成像检测[J].无损检测,2017,39(3):30-34.
备注/Memo
收稿日期: 2019-03-12 *通信作者: 黄永华(1980—),男,讲师,研究方向为测控技术、数字图像处理.
*基金项目: 福建省自然科学基金资助项目(2018J01557); 福建省激光精密加工工程技术研究中心开放基金资助项目(20150402); 福建省中青年教师教育科研项目(JAT170522)