北京航空航天大学学报 ›› 2011, Vol. 37 ›› Issue (6): 673-679.

• 论文 • 上一篇    下一篇

航空铝合金针孔缺陷自动分级的图像处理方法

吴鑫1, 齐铂金2, 张健合3   

  1. 1. 北京交通大学 机械与电子控制工程学院, 北京 100044;
    2. 北京航空航天大学 机械工程及自动化学院, 北京 100191;
    3. 北京航空材料研究院 无损检测室, 北京 100095
  • 收稿日期:2010-03-30 出版日期:2011-06-30 发布日期:2011-07-05
  • 作者简介:吴 鑫(1972-),男,吉林省吉林市人,讲师,xwu@bjtu.edu.cn.

Methods of image processing for automatic grading of porosity defects in aeronautical alloy

Wu Xin1, Qi Bojin2, Zhang Jianhe3   

  1. 1. School of Mechanical, Electronic and Control Engineering , Beijing Jiaotong University, Beijing 100044, China;
    2. School of Mechanical Engineering and Automation, Beijing University of Aeronautics and Astronautics, Beijing 100191, China;
    3. Non Destructive Testing Laboratory, Beijing Institute of Aeronautical Materials, Beijing 100095, China
  • Received:2010-03-30 Online:2011-06-30 Published:2011-07-05

摘要: 针对目前航空铝合金铸件针孔缺陷人工分级的缺点,用X射线照相获取的图像,采用一种计算机图像处理和模式识别的方法进行针孔缺陷自动分级,并主要对图像处理的算法进行了研究.根据针孔缺陷图像的灰度分布特点,采用小波分析的方法滤除低频干扰信息,保留针孔的高频信息,再经过区域分割提取针孔区域,进而提取单个针孔的尺寸特征,然后进行宏观统计和分析,通过对1~8级标准图片的统计特征进行神经网络的训练,实现了针孔的自动分级.实验结果表明,这种图像处理方法有较好的适应性.

Abstract: Aiming at the disadvantage of manual grading of porosity in aeronautical alloy cast currently, a method of automatic grading by image processing and pattern recognition of computer to images got by X-ray radiography was put forward, and the methods of image processing and pattern recognition were mainly studied. According to the characteristics of gray distribution in typical porosity image, an algorithm of wavelet was taken to filter the disturbance of low frequency, which can remain the information of high frequency. Then segmentation was taken to pick up the porosity region and dimension characteristics of single porosity further, and macroscopic statistics and analysis was carried out. Neural network training was adopted by standard images from first level to eighth level, and finally the automatic grading of porosity was realized. The experimental results show that the method of image processing has good adaptability.

中图分类号: 


版权所有 © 《北京航空航天大学学报》编辑部
通讯地址:北京市海淀区学院路37号 北京航空航天大学学报编辑部 邮编:100191 E-mail:jbuaa@buaa.edu.cn
本系统由北京玛格泰克科技发展有限公司设计开发