中文摘要：

针对目前生产线上中药口服液中可见异物检测系统存在的速度慢、精度低等问题,设计了一种可同时检测多个药瓶的基于机器视觉的异物快速检测系统;介绍了成像系统和机械系统;提出了一种基于边缘轮廓的图像配准算法,有效地消除了药瓶晃动的影响;提出了一种逐点阈值分割图像的二值化方法,能得到质量较好的二值化图像;实验结果表明,该系统能快速、有效地检测出可见异物,能较好地满足在线检测需求.

英文摘要：

Owing to limitations such as low speed and low accuracy in the detection system for visible foreign suhstances in oral liquids of traditional Chinese medicine （TCM） on assembly line. a high--speed machine--vision hosed system which can simultaneously detect multiple bottles of oral liquids is designed. The imaging system and the mechanical system arc introduced. To effectively eliminate the sloshing of the bottles, an image registration method based on edge extraction is proposed. And a binarization method which implements point-- by point segmentationon images is proposed to obtain high--quality binary images. The experimental results show that the system can effectively de tect foreign substances in oral liquids of TCM with high--speed, and meet the requirements of practical assembly lines.