Deep learning based visual pollutant detection system

Date
2020
Research Supervisor
Editor
Journal Title
Volume
Issue
Journal Title
Journal ISSN
Volume Title
Abstract
Most of us are familiar with mainly four types of pollutions. They are 1) air pollution 2) water pollution 3) sound pollution and 4) soil pollution. But a very few of us know about ‘visual pollution’ since it is relatively a new concept. But it is gaining people’s attention slowly. Determining which object should be considered as a visual pollutant is quite a challenge because an object which is visually disturbing to one person may not be visually disturbing to another person. In this study, there are four classes 1) billboards and signage 2) telephone and communication wires 3) network towers and 4) street litter. A deep learning model has been used to detect visual pollutants in an image that was trained and tested on images that we collected using the Google search engine. Our study has a lot of applications in real life like image and video analysis, visual pollutant management, visual pollution index generation, server deployment, and many more. Our study suggests that higher levels of accuracy can be achieved by increasing the size of the dataset.
Description
Keywords
TECHNOLOGY::Electrical engineering, electronics and photonics::Electrical engineering
Citation
Department Name
Electrical and Computer Engineering
Publisher
North South University
Printed Thesis
DOI
ISSN
ISBN