Segregation of Solid Municipal Waste Using Machine Learning

As the world's population is growing, so is the generated waste. To process the trash efficiently, we must separate recyclable and organic waste from non-biodegradable waste. This paper discusses an approach based on machine learning, which classifies the images of garbage into two types, Organic and Recyclable. The algorithm uses a Convolutional Neural Network (CNN), a well-known deep-learning technique for models based on data sets consisting of images and pictures. The proposed model in this paper is DenseNet-169, which contains a total of 169 layers. Data augmentation and normalization are carried out to increase the number of training images and variable sizes of the images. Results were obtained by changing the train and test split, showing different accuracies corresponding to them. The results show that the model can accurately classify unknown images. The highest achieved accuracy was about 98.630%, with the proposed model.