| 浮法玻璃缺陷在线识别技术的研究 |
| |
| 引用本文： | 刘怀广,陈幼平,谢经明,彭向前.浮法玻璃缺陷在线识别技术的研究[J].小型微型计算机系统,2011,32(4). |
| |
| 作者姓名： | 刘怀广 陈幼平 谢经明 彭向前 |
| |
| 作者单位： | 华中科技大学机械科学与工程学院,湖北,武汉,430074 |
| |
| 基金项目： | 湖北省科技攻关重点项目(20081032112)资助 |
| |
| 摘 要： | 表面缺陷是影响浮法玻璃质量的主要因素,针对目前国内在浮法玻璃缺陷识别正确率不高的现状,本文结合玻璃缺陷低灰度的特点,利用两次一维OTSU缺陷分割方法,实现了缺陷核心的有效分割.然后根据不同缺陷核心灰度分部的特点,提出12种具有统计性的特征,利用改进的神经网络的非线性映射能力实现了缺陷的正确识别.最后的实验结果证明,算法具有较好的效果.
|
| 关 键 词： | 图像处理 特征提取 缺陷识别 神经网络 |
Research on Online Recognition Technology for Float Glass Defects |
| |
| LIU Huai-guang,CHEN You-ping,XIE Jing-ming,PENG Xiang-qian.Research on Online Recognition Technology for Float Glass Defects[J].Mini-micro Systems,2011,32(4). |
| |
| Authors: | LIU Huai-guang CHEN You-ping XIE Jing-ming PENG Xiang-qian |
| |
| Affiliation: | LIU Huai-guang,CHEN You-ping,XIE Jing-ming,PENG Xiang-qian(College of Mechanical Science and Engineering,Huazhong University of Science and Technology,Wuhan 430074,China) |
| |
| Abstract: | Defects existing is the main factor effecting quality of float glass.As the low correctness of float glass defects in our country,a segmentation method of twice OTSU was proposed according to the low grey level of glass defects.The defect cores were separated correctly based on the method.Twelve statistical characters were obtained for defect recognition.Finally,defects were classified correctly with the nonlinear mapping ability of neural network.The final experiment results showed efficiency of the method... |
| |
| Keywords: | image processing pattern extract defect recognition neural network |
| 本文献已被 CNKI 万方数据 等数据库收录！ |
|
