节点文献

基于机器视觉的医药注射剂智能检测系统研究

A Machine-Vision-Based Intelligent Inspection System for Pharmaceutical Injections

  • 推荐 CAJ下载
  • PDF下载
  • 不支持迅雷等下载工具,请取消加速工具后下载。

【作者】 周博文王耀南葛继张辉

【Author】 ZHOU Bo-wen WANG Yao-nan GE Ji ZHANG Hui(College of Electronical and Information Engineering,Hunan University,Changsha 410082,China)

【机构】 湖南大学电气与信息工程学院

【摘要】 介绍了基于机器视觉的医药注射剂智能检测系统.设计了系统旋转式检测轮盘结构以及基于运动控制卡和数据I/O卡的电气控制系统.通过高速摄像机获取医药注射剂序列图像;采用二次差分和能量累积方法,获取注射剂内可能存在的运动小目标.提出了基于支持向量机的气泡和可见异物分类算法,准确判断出注射剂溶液内是否含有微小可见异物.实验验证了所提检测算法的有效性和装置的可行性.

【Abstract】 A machine-vision-based intelligent inspection system for pharmaceutical injections is introduced.The wheel mechanical structure for rotational detection and the electric control system based on motion card and data I/0 card are designed.A high-speed camera is applied to obtaining the pharmaceutical injection image sequence.Second-difference and energy accumulation methods are used to identify possible tiny moving targets in the injections.The algorithms for separating air bubbles and visible foreign substances based on support vector machine are proposed and it can accurately determine whether there are tiny visible foreign substances in the injections.The experiments demonstrate that the proposed detection algorithms are effective and the device is feasible.

【基金】 国家863计划资助项目(2007AA04Z244,2008AA04Z214);国家自然科学基金重点资助项目(60835004).
  • 【分类号】TP274
  • 【被引频次】53
  • 【下载频次】515
节点文献中: 

本文链接的文献网络图示:

本文的引文网络