Hybrid cnn–densenet121 architecture for intelligent waste classification in sustainable smart cities

Department of Computer and Information Engineering, College of Electronics Engineering, Ninevah University, Mosul, Iraq

^* Corresponding author: This email address is being protected from spambots. You need JavaScript enabled to view it.

Published online: 13 May 2026

Abstract

The growing volume of waste produced around the world has posed serious problems for recycling and environmental sustainability. Correct waste classification is a key to successful waste recycling, and the current type of systems remains dependent on manual classification, which is time-consuming and prone to errors. Thus, waste classification automation has become a significant field of study. This paper proposes a deep-learning hybrid model for waste classification. The proposed system combines a four-layer CNN and a DenseNet121 model to obtain both detailed visual features and higher-level patterns of waste images. The TrashNet dataset is used to train and test the model and consists of six classes of recyclable waste. The findings indicate that the proposed hybrid model has better performance compared to a single model system, as combining the proposed CNN with DensNet121 enabled an efficient waste classification. This research supports the development of intelligent recycling systems that enhance the sustainability and efficiency of waste management practices.