基于机器视觉的口服液智能灯检系统的研究

A Research of Automatic Liquor Detection System Based on Machine Vision

针对中药口服液的检测,本文设计了一种智能灯检系统。使用帧差法、自适应阈值分割、图像去噪和求重心偏移来识别出杂质,并用踢瓶装置将不合格产品踢除。本设计方案已在国内某些知名医药企业进行了试用,结果表明机器稳定性好、精度高、速度快。

This paper proposes a new design of automatic Liquor detection system. Through comparing the difference of two adjacent pictures, using image segmentation by self-adapting threshold, wiping off image noise, calculating the offsets of the barycenters, im- purities arc detected. The unqualified liquor products be rejected by the bottle kick device. The machine based on this design has been tested in some famous medicine companies, the result indicates that the system is stable, reliable and efficient.