Automatic dry waste classification for recycling purpose
There has been a serious increment in solid waste in the past decades due to rapid urbanization and industrialization. Therefore, it becomes a big issue and challenges which need to have a great concern, as accumulation of solid waste would result in environmental pollution. Recycling is a method wh...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Conference or Workshop Item |
| Published: |
2022
|
| Subjects: | |
| Online Access: | http://eprints.utm.my/id/eprint/101441/ https://alife-robotics.co.jp/members2022/icarob/index.html |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Institution: | Universiti Teknologi Malaysia |
| Summary: | There has been a serious increment in solid waste in the past decades due to rapid urbanization and industrialization. Therefore, it becomes a big issue and challenges which need to have a great concern, as accumulation of solid waste would result in environmental pollution. Recycling is a method which has been prominent in order to deal with the problems, as it is assumed to be economically and environmentally beneficial. It is important to have a wide number of intelligent waste management system and several methods to overcome this challenge. This paper explores the application of image processing techniques in recyclable variety type of dry waste. An automated vision-based recognition system is modelled on image analysis which involves image acquisition, feature extraction, and classification. In this study, an intelligent waste material classification system is proposed to extract 11 features from each dry waste image. There are 4 classifiers, Quadratic Support Vector Machine, Cubic Support Vector Machine, Fine K-Nearest Neighbor and Weighted K-Nearest Neighbor, were used to classify the waste into different type such as bottle, box, crumble, flat, cup, food container and tin. A Cubic Support Vector Machine (C-SVM) classifier led to promising results with accuracy of training and testing, 83.3% and 81.43%, respectively. The performance of C-SVM classifier is considerably good which provides consistent performance and faster computation time. Further classification process is improved by utilization of Speeded-Up Robust Features (SURF) method with some limitations such as longer response and computation time. |
|---|