Publication

Abstract : Pollution induced by waste is one of the major problems across the globe for a long time. During the difficult natural calamities, it worsens the situation due to the improper way of disposing of the materials subsequently causes danger to human life. This also creates a risk for sanitation workers during the process of waste collection, sorting, and recycling. So there is a need for adapting the technology in real-time for proper waste segregation and management. In this paper, a model is proposed for the classification of waste objects according to their materials which can help in the automatic recycling process. The proposed approach uses the State-of-art deep learning architecture to construct the classifier. The architecture focusing on minimizing the structural complexity and maintain better classifier accuracy. The classifier uses the features extracted automatically from the Convolutional Neural Networks (CNN) to build a model. This model classifies the waste into plastic, paper, cardboard, trash, glass, and metal with an accuracy of 87%.