医用药剂中可见异物自动化检测研究  

Medical Pharmacy Visible Foreign Matter in the Automated Inspection

作　　者：欧青[1] 任艳霞 任艳阳[1] 赵文静[1] 

机构地区：[1]中南大学物理与电子学院,长沙410083 [2]国核工程有限公司,上海200233

出　　处：《电脑与信息技术》2012年第3期24-26,37,共4页Computer and Information Technology

摘　　要：针对国内人工药品检测的低效率、低准确率的不足,文章介绍了一套根据国家行业标准设计的基于智能机器视觉的可见异物自动化检测系统。该系统主要包括机械传动模块、电气控制模块、图像识别与处理模块。根据异物运动连续性和噪声运动无序性等特点,提出了改进的三次差分和能量累积算法来提取运动杂质,然后采用基于基于SIFT特征匹配和最邻近匹配的目标跟踪算法对杂质进行跟踪,获取杂质的大小和数目并分析出杂质总类,最后检测出杂质情况。最后把检测结果与国家标准进行比对,判断出该药剂是否合格。测试表明,系统的检测准确率达90%以上,满足企业要求,基本能替代人工检测。According to the national standards of production,our team design a automatic impurity detecting system based on intelligent visual technology, against the low efficiency and low accuracy of domestic artificial detection. The system mainly includes mechanical transmission module, electrical control system module and image processing module. According to the feature：continuity of impurities movement and discontinuity of noise movement, using the third difference with the energy accumulation algorithm detects particles, then use the SIFT features and MeanShifi algorithm to track impurities, finally detect the extracted impurities, and find out whether the product is qualified. Experimental results show that the system accuracy rate can reach 90%. It has meet requirements of most enterprise, and in most cases can replace the original manual testing.

关 键 词：机器视觉 可见异物检测 序列图像三次差分 异物跟踪 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]