AI-Powered Waste Management: A Vision-Based Plastic Waste Classifier Using OpenCV

The improper disposal and inefficient sorting of plastic waste pose a significant
environmental threat. In this paper, we propose an AI-powered plastic waste classification
system using computer vision techniques with OpenCV. The proposed system captures
images of waste materials, preprocesses them using filters and contour detection, and then
classifies the type of plastic using a lightweight machine learning algorithm. This real-time
classification aids in improving the recycling process by automating the waste sorting
stage. The model is trained on a custom dataset of commonly used plastic types such as
PET and HDPE. Our system is intended to be integrated with smart bins and waste
management units to promote sustainable practices.