Abstract：A new detection method and its equipment for defects of toughened glass insulators are described.The statistical law of distribution of glass insulator defects and bands of scattered light are analyzed.Based on machine vision technology,a large area ring LED light source is used in the rotary control platform.Canny operator image segmentation and neural network technology are presented in this research in order to overcome the difficulties in defect detections caused by complex shapes of toughened glass insulators and diversity of the defects.The experiment shows this method and equipment can meet the needs of industrial automation production.
吴洪潭,石成龙,孙悦. 钢化玻璃绝缘子缺陷检测方法及装置研究[J]. 计量学报, 2012, 33(3): 203-205.
WU Hong-tan,SHI Cheng-long,SUN Yue. Development of the Method and Equipment of Defect Detection for Toughened Glass Insulator. Acta Metrologica Sinica, 2012, 33(3): 203-205.
［1］Sunil K K.Lighting design for machine vision application［J］.Image and Vision Computer,2006,24(7):720-726.
［2］Nicola B M, Ltze E, Peirs A. Non-destruction measurement of bitter pit in apple fruit using NIR hyperspectral imaging［J］. Postharvest Biology Technology,2006,40(1):1-6.
［3］Paul O L.Machine vision for feedback control in a steel rolling mill［J］. Computer in Industry, 2005, 56(4):997-1004.
［4］Su Z W, Tian G Y, Gao C H. A machine vision system for on-line removal of contaminants in wool［J］. Mechatronics, 2006, 16(3):243-247.
［5］冈萨雷斯. 数字图像处理［M］. 北京:电子工业出版社, 2004.
［6］侯卫萍,王立海.数学形态学与Canny算子在木材腐朽图像特征提取中的应用［J］. 森林工程, 2011, 27(2):27-29.
［7］朱虹. 数字图像处理基础［M］. 北京:科学出版社, 2006.
［8］白兵乾. 基于颜色和圆形度的交通禁令标志提取［J］. 电子设计工程, 2011, 19(6):150-153.
［9］李迪,陈向,续志军,等. 自组织递归区间二型模糊神经网络在动态时变系统辨识中的应用［J］. 光学精密工程, 2011, 19(6):1406-1413.
［10］Psaltis D, Sideris A, Yamamura A. A multilayered neural network controller［J］. IEEE Control Systems Magazine, 1988, 8(2):17-21.
［11］徐丽娜.神经网络控制［M］, 北京:电子工业出版社, 2003.
2012, Vol. 33 
