A Fully Automated Scrap Recognition Method Based on Convolutional Neural Networks
Tian Tan, Yuan Zhou, Mingchao Fang · 2024
As global scrap management challenges intensify, traditional recycling systems encounter inefficiencies and resource loss. The application of artificial intelligence (AI) technologies presents new opportunities to optimize the recycling process. This paper explores the use of AI in recycling, focusing on intelligent scrap classification and metal content detection, thereby enhancing recycling efficiency and reducing environmental impact. Additionally, this paper proposes a fully automated scrap recognition method. Experimental results show that this method was tested on a dataset of $\mathbf{5 0, 0 0 0 0}$ images covering $\mathbf{2 6 0}$ types of scrap, achieving a top- 1 recognition rate of eighty-five point three percent and a top- 3 recognition rate of ninety-seven point three percent.