학술논문
Object Detection Based Management System of Solid Waste Using Artificial Intelligence Techniques
Document Type
Conference
Author
Source
2022 IEEE 13th Annual Ubiquitous Computing, Electronics & Mobile Communication Conference (UEMCON) Ubiquitous Computing, Electronics & Mobile Communication Conference (UEMCON), 2022 IEEE 13th Annual. :0019-0023 Oct, 2022
Subject
Language
Abstract
Object detection-based waste management system remains a challenge for Bangladesh because of the inadequate diversity of data. Solid waste management is a grave concern for Bangladesh. By 2025 waste generation per capita will be 0.75 kg/capita/day, and the total amount of waste will reach 21.07 million tons per year. Our proposed deep learning-based waste detection approach is to detect and locate solid waste in real-time images in the context of Bangladesh. One of the main concerns was to train our model with a proper dataset that could detect many items accurately. So we collected our dataset from several open sources, and one-third of the images used in this research are our own collected images, and we fully annotated these data. Our system can detect 12 types of waste like paper, plastic, polythene, glass, metal, bio, e-waste, etc. Our best classification accuracy rate is 73%, and the F1 score is 0.729.