Waste Classification using Convolutional Neural Network

Recycling of waste from households and industries, is one of the methods that has been proposed to reduce the ever-increasing pressure on landfills. Different types of waste types warrant different management techniques and hence, proper waste segregation according to its types is essential to facilitate proper recycling. Current existing segregation method still relies on manual hand-picking process. In this paper, a method; based on deep learning and computer vision concepts, to classify wastes using their images into six different waste types (glass, metal, paper, plastic, cardboard and others) has been proposed. Multiple-layered Convolutional Neural Network (CNN) model, specifically the well-known Inception-v3 model has been used for classification of waste, with trained dataset obtained from online sources. High classification accuracy of 92.5% is achievable using the proposed method. It is envisaged that the proposed waste classification method would pave the way for the automation of waste segregation with reduced human involvement and therefore, helps with the waste recycling efforts.