基于机器视觉的垃圾分拣机器人设计与实现  

Design and Implementation of Garbage Sorting Robot Based on Machine Vision

作　　者：魏铭辰 李悦[1] 毛怡琳 房泳逸 李一鹏 Wei Ming-chen;Li Yue;Mao Yi-lin;Fang Yong-yi;Li Yi-peng(Changchun Normal University,Jilin Changchun 130000)

机构地区：[1]长春师范大学,吉林长春130000

出　　处：《内燃机与配件》2023年第23期42-44,共3页Internal Combustion Engine & Parts

基　　金：2022年国家级大学生创新创业训练项目(202210205022)

摘　　要：垃圾分类回收是实现资源循环利用至关重要的一环,对此设计了一种基于OpenMV机器视觉的垃圾分拣机器人。针对可回收垃圾、不可回收垃圾、有害垃圾选出特例作为训练集,利用EDGE IMPULSE的在线网站训练出神经网络模型,识别分类成功率可达97%以上。分拣机器人支持在线自动和手机APP控制两种模式,通过传送带传送,由舵机驱动机械臂投放至指定分类的垃圾桶。后由垃圾运输车自动运输垃圾,节省人力。Garbage sorting and recycling is an important part to realize the recycling of resources.A garbage sorting robot based on OpenMV machine vision was designed.Special cases of recyclable waste,non-recyclable waste and hazardous waste were selected as training sets,and EDGE IMPULSE's online website was used to train the neural network model with a recognition and classification success rate of more than 97%.The sorting robot supports two modes:online automatic and mobile APP control.It is transmitted through the conveyor belt and put into the designated trash can by the mechanical arm driven by the steering engine.After the garbage truck automatically transport garbage,save manpower.

关 键 词：OpenMV 机器视觉 神经网络模型 垃圾分类 

分 类 号：TP242.2[自动化与计算机技术—检测技术与自动化装置]