视觉引导的机器人条烟分拣系统设计与实现

Design and Implementation of Robot Cigarette Sorting System Based on Vision

目的 为了实现对多品种条烟高效率分拣,开发基于视觉的工业机器人条烟分拣系统平台。方法 系统采用LABVIEW图像采集及处理双进程执行系统,并基于LABVIEW进行二次开发的图像模板匹配算法。模板匹配算法得到的数据,通过与工业机器人之间的数据库对比来实现对条烟的识别和分拣,拓展机器人的应用范围。结果 4种任意放置条烟进行分拣,识别时间在100ms之内,条烟识别准确率为100%,条烟分拣速度根据传送带设置为0~4 m/s可调。结论 视觉引导的机器人条烟分拣系统软硬件设计正确,条烟识别准确,条烟分拣效率高。

The work aims to develop an industrial robot cigarette sorting system platform based on vision to sort various kinds of cigarettes efficiently. The double process execution system of LABVIEW image acquisition and processing was adopted. The image template matching algorithm of the second development was conducted based on LABVIEW. The data obtained from the template matching algorithm can be compared to the database of industrial robots to realize the recognition and sorting of cigarettes, which expanded the application scope of robots. Four kinds of cigarettes were randomly placed for sorting, the recognition time was within 100 ms, and the recognition accuracy was 100%, the cigarette sorting speed can be adjusted according to the conveyor belt setting of 0-4 m/s. The practice shows that the software and hardware design of the system is correct, the cigarette identification is accurate, and the cigarette sorting efficiency is high.