Developing a waste segregation assistant (mobile application) for households in Ghana
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
As the slowly continues to deteriorate, many organizations and sovereign states are responsible for innovating sustainable measures to reduce the production of harmful substances and repurpose materials by adding value to them. Ghana is a rapidly developing country. With this development comes a large amount of waste production, the majority of which ends up in the landfills, subsequently destroying the soil and releasing harmful gases into the air due to its poor waste management infrastructure. This project focuses on using deep learning and mobile application technology to take advantage of its massive privatized informal waste collection sector to redirect waste from landfills to treatment plants, thereby addressing specific Using transfer learning with a pretrained Keras model, a waste classification model with a 91% test accuracy was developed. This enabled the average household user to sort their waste visually. Results of the classification analysis are shared with waste collectors and recyclers registered on the mobile application to enable them to engage in the specialized collection. This approach would sensitize household users on the amount of waste they generated daily and prevent the contamination of waste items when transported to treatment plants because waste collectors would be collecting one type of waste from different locations at once.