| 摘 要: 医疗器械产品生产中的笔杆表面缺陷是不可避免的问题，基于机器视觉的自动检测方法可以克服传统人工检测效率低、漏检及误检率高等问题。在分析笔杆结构和缺陷的基础上，文章重点研究笔杆边缘直线拟合、缺陷灰度值差异、图像边缘平滑和稳定等检测方法；通过实验证明，该方法准确率可达到98.8%，每个笔杆的检测时间为8.3 s，相较于人工检测，明显提高了检测精度和速度，可以满足对笔杆实时自动缺陷检测的要求。 |
| 关键词: 机器视觉；笔杆表面缺陷检测；直线拟合；边缘平滑；尺寸测量 |
|
中图分类号: TP311.5
文献标识码: A
|
|
| A Surface Defect Detection Method for Penholder based on Machine Vision |
|
WANG Tieyu, CAO Xin
|
(Dalian Neusoft University of Information, Dalian 116023, China)
tieyu.wang@fitow.com; caoxin@neusoft.edu.cn
|
| Abstract: The surface defect detection of penholder is inevitable in producing medical devices. Automatic detection method based on machine vision can overcome problems of low efficiency, high rate of missed and false detection of traditional manual detection. Based on the analysis of the penholder structure and the defects, this paper focuses on detection methods, such as the straight line fitting of the penholder edge, the difference of the defect gray value, the smoothness and stability of the image edge, etc. Experiments show that the accuracy of the proposed method can reach 98.8%, and the detection time for each penholder is 8.3 seconds. Compared with manual detection, this algorithm significantly improves the detection accuracy and speed, and can meet the requirements of real-time automatic defect detection of penholder. |
| Keywords: machine vision; surface defect detection for penholder; straight line fitting; smooth edge; dimensional measurement |
