Segmentation and evaluation of diamond abrasive grains based on K-Means clustering and convex hull detection
-
摘要: 金刚石工具广泛应用于磨削、线锯、研磨等领域，其表面的磨粒特征是影响加工结果与工具性能的重要因素。针对具有复杂背景信息的磨粒图像，提出一种基于K-means聚类与凸包检测的磨粒分割方法，结合二值化、形态学处理、主轮廓提取等相关操作实现磨粒的提取与分割；最终提出磨粒轮廓面积精度ηCAA、磨粒位置误差θPE和磨粒数量召回率σQR 3个相关指标来评价分割效果。结果表明：金刚石磨粒的平均轮廓面积精度为95.01%，平均位置误差仅为2.93%，平均数量召回率为98.30%，证明了该方法的准确性。Abstract: Diamond tools are widely used in grinding, wire sawing and other fields. The characteristics of abrasive particles on the surface are an important factor affecting the machining results and tool performance. To process abrasive grain images with complex background information, this paper proposed an abrasive grain segmentation method based on K-means clustering and convex hull detection, which combines binarization, morphological processing, main contour extraction and other related operations to achieve abrasive grain extraction. Finally, three related indicators, including abrasive grain contour area accuracy ηCAA, abrasive grain position error θPE, and abrasive grain quantity recall rate σQR, were proposed to evaluate the segmentation effect. The results show that the average contour area precision is 98.30%, the average position error is only 2.93%, and the average number recall rate is 95.91%, which proves the accuracy of the method.
-
图 8 凸包检测结果及黏连磨粒分割结果示意图
Figure 8. Schematic diagram of convex hull detection results and adhesion abrasive particle segmentation results
图 12 使用前磨粒图像基于KM-CHD方法的分割面积以及面积精度
Figure 12. Segmentation area and area accuracy of unused abrasive image based on KM-CHD method
图 10 使用前磨粒图像及其KM-CHD分割结果
Figure 10. Image of abrasive grains before use and its KM-CHD segmentation results
图 11 使用后磨粒图像及其KM-CHD分割结果
Figure 11. Image of abrasive grains after use and its KM-CHD segmentation results
图 13 使用后磨粒图像基于KM-CHD方法的分割面积以及面积精度
Figure 13. Segmentation area and area accuracy of used abrasive image based on KM-CHD method
图 14 使用前磨粒图像基于KM-CHD方法与手动分割的磨粒质心图
Figure 14. The centroid of unused abrasives image segmented by the KM-CHD method and the manual method
图 16 使用后磨粒图像基于KM-CHD方法与手动分割磨粒质心图
Figure 16. The centroid of used abrasives image segmented by the KM-CHD method and the manual method
-
[1] ZHANG Y, FANG C, HUANG G, et al. Modeling and simulation of the distribution of undeformed chip thicknesses in surface grinding [J]. International Journal of Machine Tools and Manufacture,2018,127:14-27. doi: 10.1016/j.ijmachtools.2018.01.002 [2] BI G, ZHENG S, ZHOU L. Online monitoring of diamond grinding wheel wear based on linear discriminant analysis [J]. The International Journal of Advanced Manufacturing Technology,2021,115(78):2111-2124. [3] 张秀芳, 于爱兵, 贾大为, 等. 应用数字图像识别法检测金刚石磨粒的形状与粒度 [J]. 金刚石与磨料磨具工程,2007(1):47-49. doi: 10.3969/j.issn.1006-852X.2007.01.012ZHANG Xiufang, YU Aibing, JIA Dawei, et al. Measuring the shape and size of diamond grains by digital image identification method [J]. Diamond & Abrasives Engineering,2007(1):47-49. doi: 10.3969/j.issn.1006-852X.2007.01.012 [4] 李银华, 路新惠, 靳贺敏. 基于图像处理的金刚石磨粒体积计算研究 [J]. 计算机工程与设计,2009,30(18):4242-4244. doi: 10.16208/j.issn1000-7024.2009.18.001LI Yinhua, LU Xinhui, JIN Hemin. Calculation of volume for diamond grains based on image processing [J]. Computer Engineering and Design,2009,30(18):4242-4244. doi: 10.16208/j.issn1000-7024.2009.18.001 [5] 吴文艺, 崔长彩, 叶瑞芳, 等. 采用二次灰度直方图的砂轮磨粒图像阈值分割[J]. 华侨大学学报(自然科学版). 2016, 37(4): 422-426.WU Wenyi, CUI Changcai, YE Ruifang, et al. Image segmentation method using second time gray level histogram of connected component labeling of grinding wheel abrasives grains [J]. Journal of Huaqiao University (Natural Science Edition), 2016, 37(4): 422-426. [6] 杨栖凤, 崔长彩, 黄国钦. 金刚石砂轮表面二维形貌全场测量和分析 [J]. 华侨大学学报(自然科学版),2018,39(4):479-484.YANG Qifeng, CUI Changcai, HUANG Guoqin. Measurement and analysis of two-dimensional surface topography of whole grinding wheel [J]. Journal of Huaqiao University (Natural Science Edition),2018,39(4):479-484. [7] 潘秉锁, 潘文超, 刘子玉. 基于空洞卷积神经网络的金刚石图像语义分割 [J]. 金刚石与磨料磨具工程,2019,39(6):20-24. doi: 10.13394/j.cnki.jgszz.2019.6.0004PAN Bingsuo, PAN Wenchao, LIU Ziyu. Semantic segmentation of diamond images based on hollow convolutional neural networks [J]. Diamond & Abrasives Engineering,2019,39(6):20-24. doi: 10.13394/j.cnki.jgszz.2019.6.0004 [8] LIN Y, FANG C, DENG Y. Segmentation and extraction for the diamond grain image based on the gauss-polymerized enhancement [J]. Journal of physics: Conference Series, 2019,1169(1):12042. [9] LIN Y, FANG C. Study on the segmentation of abrasive grains in diamond tools [J]. International Journal of Abrasive Technology,2018,3(8):203-217. [10] PAN B, YANG Y, ZHANG Y. Extraction of diamond grain topography from diamond tool surface using 3D surface measurement coupled with image analysis [J]. Measurement,2019,133:9-13. doi: 10.1016/j.measurement.2018.10.003 [11] 赵玉康, 毕文波, 葛培琪. 电镀金刚石线锯表面磨粒分布密度的多相机视觉检测 [J]. 金刚石与磨料磨具工程,2021,41(2):64-68. doi: 10.13394/j.cnki.jgszz.2021.2.0011ZHAO Yukang, BI Wenbo, GE Peiqi. Multi-camera visual inspection of abrasives distribution density on electroplated diamond wire saw surface [J]. Diamond & Abrasives Engineering,2021,41(2):64-68. doi: 10.13394/j.cnki.jgszz.2021.2.0011 [12] KANG M, ZHANG L, TANG W. Study on three-dimensional topography modeling of the grinding wheel with image processing techniques [J]. International Journal of Mechanical Sciences,2020,167:105241. doi: 10.1016/j.ijmecsci.2019.105241 [13] 赵文昌, 李忠木. 融合改进人工蜂群和K均值聚类的图像分割 [J]. 液晶与显示,2017,32(9):726-735. doi: 10.3788/YJYXS20173209.0726ZHAO Wenchang, LI Zhongmu. Image segmentation algorithm based on improved artificial bee colony and K-mean clustering [J]. Liquid Crystal and Display,2017,32(9):726-735. doi: 10.3788/YJYXS20173209.0726 [14] 段明义, 卢印举, 张文. 一种改进的舰船合成孔径雷达图像分割方法 [J]. 太赫兹科学与电子信息学报, 2021, 19(5): 905-909.DUAN Mingyi, LU Yinju, ZHANG Wen. An improved ship synthetic aperture radar image segmentation method [J]. Journal of Terahertz Science and Electronic Information, 2021, 19(5): 905-909. [15] 彭金喜, 苏远歧, 薛笑荣. 一种小波域K-Means遥感图像分类标注算法 [J]. 软件导刊,2019,18(9):202-206.PENG Jinxi, SU Yuanqi, XUE Xiaorong. A remote sensing image semantic classification label of K-means clustering on wavelet transform [J]. Software Guide,2019,18(9):202-206. [16] 李冰, 何超. 基于背景骨架特征的粘连米粒图像分割算法 [J]. 计算机应用,2017,37(S2):198-202.LI Bing, HE Chao. Segmentation algorithm of touching rice kernels based on skeleton features of image background [J]. Computer Applications,2017,37(S2):198-202. [17] 吴忻生, 刘洋, 戚其丰. 基于凹性分析的粘连车辆分割 [J]. 计算机应用研究,2012,29(1):344-347. doi: 10.3969/j.issn.1001-3695.2012.01.095WU Xinsheng, LIU Yang, QI Qifeng. Adhered vehicles segmentation based on concavity analysis [J]. Computer Application Research,2012,29(1):344-347. doi: 10.3969/j.issn.1001-3695.2012.01.095 -
下载:
点击查看大图
计量
- 文章访问数: 370
- HTML全文浏览量: 117
- PDF下载量: 26
- 被引次数: 0
邮件订阅
RSS
