基于图像处理的智能分类垃圾桶的设计

Design of intelligent separated-waste containers based on picture processing

该项目采用OPEN-CV图像处理技术、树莓派4B开发板及卷积神经网络算法,对实现垃圾智能分类方法进行了研究并设计了一款智能分类垃圾桶。该垃圾桶包含四个子垃圾桶,由四个舵机单独控制,可以实现干垃圾、湿垃圾、有害垃圾和可回收垃圾等四种垃圾的分类回收。采用树莓派摄像头完成四种垃圾图像的采集并保存,在Pycharm中利用卷积神经网络模型训练采集到的垃圾图像,模型训练完成后,将待识别的垃圾图像在训练好的模型中分析、对比、评估、识别出该垃圾的种类并将分类信号传送到树莓派开发板中,同时树莓派开发板驱动舵机旋转,带动对应的子垃圾桶转到投放口下方完成垃圾投放,实现垃圾智能分类。

This project adopts open-CV image processing technology,raspberry PI 4B development board and convolutional neural network algorithm to study the method of realizing intelligent garbage classification and design an intelligent garbage classification bin.The trash can contains four sub-trash cans,controlled by four steering engines alone,can realize the classification and recovery of dry garbage,wet garbage,harmful garbage and recyclable garbage.With the help of four kinds of garbage raspberries pie camera,image collection and preservation,using convolution neural network model training in Pycharm garbage collected images,model training is completed,will be to identify spam image in the trained model analysis,comparison and evaluation,to identify the types of the garbage classification and signals to the raspberry pie development board,At the same time,the raspberry PI development board drives the steering gear to rotate,and drives the corresponding sub-trash can to transfer to the lower part of the garbage delivery port to complete garbage delivery and realize intelligent garbage classification.