基于机器视觉的注塑胶管品质检测的系统设计

Design of Injection Tube Quality Inspection System based on Machine Vision

以注塑胶管为研究对象,基于机器视觉检测技术,设计了注塑胶管表面黑点在线检测系统。利用CCD摄像机、面光源完成图像获取,结合VisionMaster开发的软件系统经图像预处理、特征提取等算法完成对胶管表面缺陷的自动检测。试验结果表明,该系统识别准确度达99.5%并且具有较高的检测速度,可以满足工业领域在线检测的要求。

Based on the machine vision detection technology, an online detection system for surface black spots of injection tube was designed. We use CCD camera and surface light source to complete image acquisition. With the software system developed by Vision Master, the automatic detection of the surface defects of the tube is accomplished by image preprocessing and feature extraction. The experimental results show that the system has 99.5 % recognition accuracy and high detection speed, which can meet the requirements of online detection in industry.