摘要
全自动生产线上对制动盘进行检测,在众多的无损检测方法中选择磁粉检测。简述了磁粉检测的原理以及真伪缺陷的区别,对原有的梯度法、自适应动态阈值分割算法进行分析,对其不适应制动盘检测部分进行简单说明,在此基础上选取梯度值算法、自适应动态阈值分割算法与数值分析相结合,利用MATLAB软件提取灰度值图片然后进行分析,省去专人检测环节,达到全自动化要求。实验验证该方法能够准确的识别真伪缺陷,提高了检测精度并且节约了检测时间,适用于流水线式制动盘检测。
Full automatic production line on the brake disc detection, in many non-destructive testing methods to select the magnetic particle detectiora The principle of magnetic particle inspection and the difference between the defect and the defect are described in this paper, and the original gradient method and the adaptive dynamic threshold segmentation algorithm is analyzed, it is not adapted to the brake disc test part of a simple description, Then, the gradient value algorithm, the adaptive dynamic threshold segmentation algorithm and the numerical analysis are selected and combined on the basis of this. Use MATLAB software to extract gray values and analyze which eliminates the special detection link to achieve full automation requirements. Experimental results show that the method can accurately identify the authenticity of the defect, improve the detection accuracy and save the detection time. Thus, the method is applicable for brake disc inspection on assembly line.
出处
《机械设计与制造》
北大核心
2018年第1期156-158,共3页
Machinery Design & Manufacture
基金
长春市科技计划项目(2014220)
吉林省省级经济结构战略调整引导资金专项项目(2014Y131)
关键词
制动盘
磁粉检测
图像处理
缺陷识别
自动化
Brake Disc
Magnetic Particle Inspection
Image Processing
Defect Recognition
Automation