摘要
介绍了钢球表面缺陷的常见类型,设计了基于电容传感器的大尺寸钢球表面缺陷检测系统,并对钢球沿子午线展开时的测量结果进行了分析,通过数据分布的概率密度函数直观地展示了合格钢球与缺陷钢球的测量值分布特征。根据钢球的标准参数变化提出一种动态阈值缺陷检测算法,该算法采用整体结合局部的特征值,利用缺陷表面与正常表面的测值差异对表面缺陷进行检测,并以直径18.2562 mm的不锈钢钢球为例,对表面磨损情况进行了分析验证,有效分离了缺陷数据与正常数据。
Common types of steel ball surface defects is introduced, the detection system based on the capacitive sensor for large length scale steel ball surface defects is designed, and the measurement results while the steel ball unfolding along the meridian are analyzed. Through the probability density function of the data distribution, the distribution characteristic of measurement values between the qualified steel balls and defective ones is directly showed. According to the standard parameters change of steel balls, a dynamic threshold detection algorithm for defects is presented. By the whole features value combining with local ones, the algorithm uses differences of measurements between defective surface and normal surface to detect the surface defect. Then taking diameter 18. 2562 mm stainless steel balls for example, the surface wear situation is analyzed and tested, and the data with defective feature and data with normal feature are effectively separated.
出处
《传感器与微系统》
CSCD
北大核心
2012年第8期144-146,共3页
Transducer and Microsystem Technologies
基金
山东省教育厅资助项目(J10LD22)
山东省博士基金资助项目(BS2010ZZ001)
山东省高等学校一般国内访问学者资助项目
关键词
电容传感器
大尺寸钢球
表面缺陷
阈值检测
capacitive sensor
large length scale steel ball
surface defect
threshold detection
