Automated Garbage Classification and Sorting System using Machine Learning

 




 

Gan, Ji Sheng (2020) Automated Garbage Classification and Sorting System using Machine Learning. Final Year Project (Bachelor), Tunku Abdul Rahman University College.

[img] Text
Gan Ji Sheng Full Text.pdf
Restricted to Registered users only

Download (6MB)

Abstract

Machine learning is a study of algorithm that enable the computers to learn how to do a given task without being pre-programmed to do so. It basically emulates the way human learns and get better at tasks through learning. In today’s world, the increasing amount of garbage created over the years have become a major concern. An efficient waste management is needed for a well-managed recycling and able to provide a greener and healthier environment. To solve this problem, YOLO is proposed in this project. YOLO is an object detection algorithm which is convolutional neural network based deep learning. It has the capabilities of detecting multiple object at once in real time. In this project, it involves garbage with different shapes. Thus if the variations are too high, it will cause a lot of false detection. Besides that, YOLO model requires decent amount of computation power to run and a low power device like raspberry pi will have a hard time to run it. This is why the main objectives in this project is to be able to identify six classes of garbage which is cardboard, glass, metal, paper, plastic and trash using YOLO model with accuracy of 70% and above in garbage classification. And lastly, it is to be able to run the model on raspberry pi with frame per second (FPS) of 2 and above. The use of pre-trained YOLO and fine tune method are utilized to achieve the objectives. Throughout this project, research on the field of deep learning, convolutional neural network and YOLO through previous works are presented and discussed. The research methodology as well as the results obtained are being justified and discussed. In the end, all the objectives are able to be achieved in this project.

Item Type: Final Year Project
Subjects: Science > Computer Science
Technology > Electrical engineering. Electronics engineering
Technology > Technology (General) > Automation
Faculties: Faculty of Engineering and Technology > Bachelor of Mechatronics Engineering with Honours
Depositing User: Library Staff
Date Deposited: 29 Apr 2020 16:40
Last Modified: 02 Oct 2020 07:45
URI: https://eprints.tarc.edu.my/id/eprint/14496