Waste Classification for Smart Recycling Systems: Image-Based Deep Learning with ResNet-50 CNN

View at publisher Summarise with VIRAT
Authors: A. Gayathri, S. Christy, C. Sivasankar, A. Tamilarasan
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Precise waste type classification is crucial for both environmental conservation and effective waste management. Traditional garbage classification methods are labour-intensive and liable to human error. By employing an optimised ResNet-50 Convolutional Neural Network (CNN) structure, this research paper suggested an image-based deep learning model for automated garbage type classification. With the TrashNet dataset, the model identifies trash into classes such as cardboard, organic, trash, plastic, metal, glass, and paper. These labeled data were used to train and test the algorithm, and the algorithm demonstrated very high accuracy and robustness in real-world scenarios. ANOVA results showed significant improvements over baseline models, and SPSS-based statistics verified the significance of the model performance. The proposed method is well suited for implementation in intelligent processing systems, as it outperforms standard CNN architectures in accuracy and computation speed.

Fields of Study & Subjects

Advanced Neural Network Applications Municipal Solid Waste Management Waste Management and Recycling

Publication details

Journal / source
CRC Press eBooks
Pages
160–167
Publication date
31 March 2026
Language
EN
Document type
Book chapter