基于树莓派的智能分类垃圾投放系统  

Intelligent Classification of Garbage Disposal System Based on Raspberry Pi

作　　者：刘少儒 雷宇 徐寅林[1] 

机构地区：[1]南京师范大学计算机与电子信息学院/人工智能学院,江苏南京210023

出　　处：《工业控制计算机》2022年第1期43-46,共4页Industrial Control Computer

摘　　要：基于树莓派设计了一种新型智能分类垃圾投放系统。采用MobileNet V2算法并进行迁移学习获得最终分类模型。为解决数据集类别不均衡问题,采用先细分后粗分两级分类方法以获得高分类正确率。以树莓派作为图像采集及分类处理器,用单片机控制相关机电部分工作,最终实现垃圾的智能检测及投放。实验结果表明,该投放系统进行垃圾分类达到了94.6%以上的准确率。Based on the Raspberry Pi,a new type of intelligent classification of garbage disposal system is designed in this paper.Use MobileNet V2 algorithm and transfer learning to obtain the final classification model.In order to solve the problem of the imbalance of the data set,a two-level classification method is adopted to obtain a high classification accuracy rate.Raspberry Pi is used as an image acquisition and classification processor,and a single-chip microcomputer is used to control the work of related electromechanical parts,and realize the intelligent detection and disposal of garbage.The experimental results show that the classification of garbage by the system has reached an accuracy rate of over 94.6%.

关 键 词：垃圾投放 MobileNet V2 迁移学习 两级分类 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术] TP18[自动化与计算机技术—计算机科学与技术] X799.3[环境科学与工程—环境工程]