Research on Intelligent Garbage Classification Device Based on the Fusion Technology of Differential Classification and Artificial Intelligence Recognition

Abstract

To solve the problem of the "last mile" in garbage classification, this study designs an intelligent garbage classification device based on artificial intelligence, machine vision and automation technologies. By constructing a multi-modal garbage identification model (YOLOv5s) and integrating it with a differential sorting mechanism, the functions of garbage type identification, sorting and compression are achieved. The experimental results show that the device's recognition accuracy rate is ≥98%, the single-piece processing cycle is ≤300ms, the compression rate of recyclable waste is ≥70%, the operating cost is reduced by 60%, and the sorting rate of hazardous waste is increased to 99%. This device overcomes the bottlenecks of traditional classification efficiency and accuracy, providing core infrastructure support for the construction of "waste-free cities" and contributing to the achievement of the "dual carbon" goals.