The solid waste generated has been increasing rapidly due to the growth in population and consumption. This is a major issue, particularly in cities. Waste segregation is currently mostly done by hand, which is time consuming and energy demanding, and can be inaccurate when various kinds of waste are mixed. This calls for more efficient techniques. The use of machine learning and deep learning is making the classification of waste more efficient. These methods can assist in automatically classifying different waste items from images. But most of these systems have been developed for simple problems, such as classifying only two types of waste, and also work well only for a good image. This makes them less suitable to use in real scenarios when waste is not well categorised or images are not ideal. In this study, various machine learning and deep learning techniques are discussed, particularly convolutional neural networks and networks such as VGG. Their effectiveness is evaluated in terms of their classification accuracy for various types of waste and under different circumstances. The research also identifies some issues and suggests how they can be addressed. The overall objective is to learn how to make these technologies more applicable to waste management.
Keywords
Waste Classification, Machine Learning, Deep Learning, CNN, VGG, Waste Management
Conclusion
Waste management is really important these days because the world is generating a lot of waste. The old ways of sorting waste are not very good. Can lead to mistakes. That is why using technologies like machine learning and deep learning can help make waste classification better and more accurate.
In this study we looked at machine learning and deep learning methods for sorting waste based on images. We paid attention to convolutional neural networks and transfer learning methods. The results show that deep learning models, those that use VGG16 can do a good job by learning from waste images.
The model we used can classify types of waste with good accuracy. Both the training and validation results show that this approach is effective and can be used in life. This means that such systems can reduce the need for work and make waste sorting and recycling better.
However, we also found some limitations. The models performance can be affected by things like image quality, background noise and waste objects that look similar. In life these issues can reduce accuracy. Using an more diverse dataset can also make the model better and more reliable.
In the future we can explore advanced models to improve classification results. We can also work on increasing the number of waste categories and making the system more practical for life. If we can integrate these models with waste management system’s we can create more efficient and scalable solutions. Overall, this area has a lot of potential. can help with environmental management.
References
[1]G. Thung and M. Yang, “Classification of trash for recyclability status,” in Proc. Stanford Univ. CS229 Project Rep., Stanford, CA, USA, 2016, pp. 1–6.
[2]Y. Wang, W. Zhao, J. Xu, and R. Hong, “Recyclable waste identification using convolutional neural networks and clustering techniques,” in Proc. IEEE Int. Conf. Image Processing (ICIP), 2020,
pp. 1–6.
[3]X. Zhang, H. Wang, and Y. Li, “Waste image classification based on transfer learning and convolutional neural networks,” Waste Management, vol. 135, pp. 150–157, 2021.
[4]A. Chakraborty, S. Raha, and T. Chakraborty, “Computer vision- based automatic sorting of non-biodegradable waste using deep learning,” International Research Journal on Advanced Engineering Hub, vol. 5, no. 1, pp. 30–36, 2025.
[5]S. Kruthika, R. Rajadevi, and D. Sathya, “Garbage classification using deep learning techniques,” International Research Journal on Advanced Engineering Hub, vol. 2, no. 12, pp. 2774–2780, 2024.
[6]A. Riduan, F. Panjaitan, S. Rizal, N. Huda, and S. D. Purnamasari, “Detection of inorganic waste using convolutional neural networks,” Journal of Information Systems and Informatics, vol. 6, no. 1, pp. 1–10, 2024.
[7]
M. Shahid and K. R. Sumana, “Deep learning-based automatic garbage classification system,” International Journal for Research in Applied Science and Engineering Technology, vol. 11, no. 6, pp. 112–118, 2023.
[8]J. C. Simbolon, R. Simbolon, and H. Hendri, “Household waste image classification using deep learning models,” Journal of Artificial Intelligence and Engineering Applications, vol. 5, no. 1, pp. 15–22, 2024.
[9]L. Zhang, K. Chen, and J. Li, “Recyclable waste image recognition based on deep learning,” Resources, Conservation and Recycling, vol. 171, pp. 105636, 2021.
[10]J. Ma, Y. Zhao, and W. Chen, “Classification of construction waste materials using deep convolutional neural networks,” Automation in Construction, vol. 122, pp. 103481, 2021.
[11]Z. Mumtaz, R. Ahmad, and Z. Khan, “Improved waste image classification using attention-based VGG architecture,” Journal of Artificial Intelligence and Smart Systems, vol. 7, no. 3, pp. 88–96, 2025.
[12]M. K. N. Zamani, M. Z. M. Zukhi, and M. M. Din, “Development of a machine learning-based waste classification system using VGG-16 CNN,” Journal of Information and Technology Management, vol. 16, no. 2, pp. 45–53, 2024.
[13]A. Krizhevsky, I. Sutskever, and G. E. Hinton, “ImageNet classification with deep convolutional neural networks,” in Advances in Neural Information Processing Systems (NeurIPS), 2012, pp. 1097–1105.
[14]K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in Proc. IEEE Conf. Computer Vision and Pattern Recognition (CVPR), 2016, pp. 770–778.
How to Cite This Paper
Anjali Singh, Ankur Chaudhary (2026). Evaluating the Effectiveness of Machine Learning Models in Waste Classification. International Journal of Computer Techniques, 13(2). ISSN: 2394-2231.
