A Household Waste Classification System Based on Edge Computing and YOLOv7-T Algorithm
Chao Huang, Heiqiang Huang, Peng Zhao · 2024
Waste classification has always been a hot topic in environmental protection and resource recycling. However, traditional waste classification models often rely on manual labor, leading to inefficiency. Therefore, this paper proposes a household waste classification system based on edge computing and YOLOv7-T algorithm. Firstly, in order to enhance the versatility of waste classification systems, a household waste classification system serving smart home services is developed, including a complete set of automatic waste disposal solutions. Secondly, to address the issues of low accuracy and poor real-time performance in waste classification, we propose the YOLOv7-T model. Compared with the original YOLOv7, the improved network reduces the parameter volume by 28.2%, achieving an average detection accuracy increase of 1.36% on our self-made dataset. The average processing time is only 9.75 seconds. Experimental results show that the developed waste classification system performs well in terms of accuracy and real-time performance.