Improved Convolutional Neural Network for Garbage Separation

Improper waste disposal is a significant environmental issue that contributes to pollution and resource depletion. Effective recycling requires accurate classification of waste materials, which is often challenging for the average citizen. This project aims to develop a Convolutional Neural Network (CNN) that classifies trash not only as recyclable or non-recyclable but into various subcategories, considering the different degrees of recyclability and appropriate disposal methods. By accurately categorizing waste such as plastics, paper, metals, glass, and hazardous items like batteries, the system enhances recycling habits and reduces environmental impact. The CNN model will be trained on a robust dataset to ensure reliable accuracy, with the target of achieving an accuracy of over 80%. This solution will enable users to quickly and easily determine the appropriate disposal method by simply taking a photo of the trash item, ultimately promoting sustainable waste management practices.