基于语音交互与图像识别的多功能垃圾分类系统设计

Design of Multifunctional Garbage Sorting System Based on Speech Interaction and Image Recognition

随着社会的进步和人们生活水平的提高,垃圾的产生量也随之增加,意味着资源消耗量不断加大,垃圾分类刻不容缓。但垃圾分类政策实施以来,分类效果不佳,依然存在分类难、效率低等问题。为此,文章提出采用计算机视觉技术、CNN卷积神经网络以及语音交互进行垃圾图像识别以实现垃圾分类。使用树莓派驱动外接摄像头采集垃圾图像,上传图像并解析返回结果,使用STM32单片机作为底层驱动核心板驱动电机,完成对单个目标的分类。

With the progress of society and the improvement of people's living standards,the amount of generated garbage has also increased,which means that resource consumption is constantly increasing,and garbage classification is urgent.However,since the implementation of the garbage classification policy,the classification effect has been poor,and there are still problems such as difficulty in classification and low efficiency.Therefore,this paper proposes to use computer vision technology,CNN convolutional neural networks,and voice interaction for garbage image recognition to achieve garbage classification.Use a Raspberry Pi to drive external camera to collect garbage images,upload images,and parse the returned results.Use the STM32 Single-Chip Microcomputer as the underlying driver core board to drive the motor,completing the classification of individual targets.