Recyclable Waste Classification Using Computer Vision And Deep Learning
Nadish Ramsurrun, Geerish Suddul, Sandhya Armoogum, Ravi Foogooa
Abstract
Recycling solid waste is an important step to reduce harmful impact such as sanitary and health problems resulting from the over use of landfills. Yet, recycling requires the sorting of solid waste, which is complex and expensive. In an attempt to ease this process, our work proposes a Deep Learning approach using computer vision to automatically identify the type of waste and classify it into five main categories: plastic, metal, paper, cardboard and glass. Our conceptual system consists of an automated recycling bin which automatically opens the lid corresponding to the type of waste identified. This work focuses mainly on the Machine Learning algorithms which can be trained for efficient identification. Pre-existing images have been used to train a minimum of 12 variants of the Convolutional Neural Network (CNN) algorithm over three classifiers: Support Vector Machine (SVM), Sigmoid and SoftMax. Our results show that VGG19 with SoftMax classifier has an accuracy of around 88%.