摘要
针对目前国内生产聚晶金刚石复合片(PDC)的企业大多数在表面缺陷检测环节严重依赖于人工检测,存在检测效率低、主观性强等问题,提出了一种基于机器视觉的检测方法替代人工检测。将倒角边缘的崩角缺陷作为检测目标,研究崩角图像的表面特征后,提出了在硬件上使用零角度环形光源突出崩角特征,在检测方式上通过阈值分割、中值滤波进行预处理,然后利用最小二乘法拟合获取倒角圆环的圆心位置和小圆半径并建立掩码,最后通过与(AND)运算提取出崩角信息,进行识别和标记。结果显示:实现对图像中的崩角缺陷自动检测和定位,并且判断标准统一。
Aiming at problem that most domestic enterprises which produce polycrystalline diamond compact （PDC） rely heavily on manual detection on surface defect detection and manual detection leads to some problems such as low detection efficiency and strong subjectivity, propose a method based on machine vision to replace manual detection. Select the chamfered edge defect as the detection target, study the surface characteristics of the edge defect image, put forward using zero angle ring light source on hardware prominent edge defect characteristics, preprocess with threshold segmentation and median filtering, and then use the [east squares fitting to obtain circle center position and the radius of the smaller circle of chamfer ring and build a mask. Finally, the edge defect information is extracted by AND operation, and then identify and mark. The results show that the automatic detection and positioning of the edge defect in the image are realized, and judgment criteria is unified.
出处
《传感器与微系统》
CSCD
2017年第7期53-56,共4页
Transducer and Microsystem Technologies
基金
国家科技支撑计划资助项目(2012BAF13B04)
华侨大学研究生科研创新能力培育计划资助项目(1511403001)
关键词
视觉检测
缺陷检测
掩码
聚晶金刚石复合片
vision detection
defect detection
mask
polycrystalline diamond compact（PDC）
