Recyclable waste classification using YOLO-based Convolutional Neural Network (CNN) / Shaharil Armin Sharizal

Sharizal, Shaharil Armin (2025) Recyclable waste classification using YOLO-based Convolutional Neural Network (CNN) / Shaharil Armin Sharizal. Degree thesis, Universiti Teknologi MARA, Terengganu.

Abstract

Effective waste sorting management is essential for recycling efficiency and environmental challenges. Traditional methods of waste sorting, based on manual labor, often lead to inaccurate and inefficient results. An image recognition system using YOLO-based CNN is proposed in this study, which is in turn driven by a cyclical research design. After comparative analysis, the model that performed best is YOLOv8. This model achieved, by and large, over 90% classification accuracy in distinguishing between plastic, paper, and glass waste types. The model is able to extract critical visual features for effective classification after training on a prepared dataset. A simple interface is developed for real-time waste detection and identification, which is useful for waste management professionals. This is an important move forward since AI is now being used, even in its infancy, in different ways to improve on traditional systems to aid and promote sustainable waste management and environmental conservation.

Metadata

Item Type: Thesis (Degree)
Creators:
Creators
Email / ID Num.
Sharizal, Shaharil Armin
2023379823
Contributors:
Contribution
Name
Email / ID Num.
Thesis advisor
Abdul Hamid, Nor Hasnul Azirah
UNSPECIFIED
Subjects: Q Science > QA Mathematics > Instruments and machines > Electronic Computers. Computer Science > Algorithms
Divisions: Universiti Teknologi MARA, Terengganu > Kuala Terengganu Campus > Faculty of Computer and Mathematical Sciences
Programme: Bachelor of Computer Science (Hons)
Keywords: YOLO, Convolutional Neural Network (CNN), Recycling Efficiency
Date: 2025
URI: https://ir.uitm.edu.my/id/eprint/115278
Edit Item
Edit Item

Download

[thumbnail of 115278.pdf] Text
115278.pdf

Download (92kB)

Digital Copy

Digital (fulltext) is available at:

Physical Copy

Physical status and holdings:
Item Status:

ID Number

115278

Indexing

Statistic

Statistic details