An Image Processing Approach to Identify Plastic Garbage

Abstract

Waste management system can become more beneficial for us if we can recycle the plastic wastes present in garbage. For recycling, we need to sort out the plastics from other wastes in the garbage. The aim of our work is to build a model which can help sorting out plastic wastes from garbage by image processing technique. Therefore, we have built a model by applying two image processing algorithms (SVM and CNN) and compared their accuracies on our image dataset. Our image dataset is a collection of raw images, captured by us from our surroundings. We mainly focused on identifying plastic wastes, hence the images that we have collected are mainly of used plastics, not a brand new piece of plastic that is never used before. After analyzing the algorithm results, we have got that, the CNN algorithm gives better results compared to SVM for our model.