Streamlining Plastic Recycling using Machine Learning-Based Image Classification

Not all plastic is recyclable, yet many consumers rely on the recycling symbol alone, leading to contamination in recycling facilities and increased landfill waste. Uncertainty about recyclability also results in unnecessary disposal, leading plastics that could have been recycled to instead contribute to long-term environmental damage. Addressing this issue requires an accessible and accurate method for classification. This study explores the potential of machine learning to identify and classify plastic waste, helping consumers make informed recycling decisions. A custom dataset of over 10,000 images was used to train deep learning models, such as VGG-16 and VGG-19. Evaluation metrics included accuracy, recall, precision, and F1-score. The best-performing model achieved an 87.8% classification accuracy, demonstrating its effectiveness in distinguishing between recyclable and non-recyclable plastics. This model was then integrated into a mobile application that enables users to take a picture of plastic waste and receive real-time classification and disposal guidance. By reducing contamination in recycling streams and improving waste sorting, this approach supports environmental sustainability. In the future, AI-driven waste classification can reduce landfill waste, plastic pollution, and resource consumption,helping mitigate the long-term environmental impact of plastic waste.