摘要
磁粉探伤的发展趋势是自动化、智能化,而工件表面状况、真伪裂纹、工况条件等使得现有的检测识别方法难以满足工件表面裂纹缺陷自动检测识别的需要。在分析工件表面荧光磁粉图像特征及裂纹缺陷特征的基础上,研究表征裂纹邻域像素空间相关度的二维直方图分布,提出基于多重分块极值的图像边缘检测算法。根据裂纹邻域像素空间相关度参数,以及裂纹缺陷的长宽比、圆形度等特征,设计了基于Fisher线性判别方法的工件裂纹识别算法。以此为基础的荧光磁粉探伤工件裂纹缺陷自动检测识别技术,应用于列车轮轴检测线实时检测,裂纹缺陷的有效检出率达99%。
Tendency of magnetic powder crack detection is automation and intelligentization. Because of exterior status, veritable or feigned crack object, site condition, etc., existing method can’ tsuccessfully automatically detect and identify cracks on the surface. Fluorescent magnetic powder image and crack image characteristics are analyzed, pixel spatial correlative degree of crack neighboring area is studied through two dimensional histogram distribution, image fringe detection based on multi-sub-area extremum is brought forward. Crack identification algorithm based on Fisher linear discrimination method is designed according to pixel spatial correlative degree of crack neighboring area, long-width ratio and round shape degree, etc. Railway Car wheel axis crack detection line equipped with this auto-detection and identification technology reaches an efficient crack detection ratio as high as 99%.
出处
《铁道技术监督》
2010年第10期6-10,共5页
Railway Quality Control
关键词
轮对
磁粉探伤
识别
图像处理
Wheel set
Magnetic Powder Detection
Identification
Image Processing
