Waste Type Classification Using Convolutional Neural Network

Abstract

The increasing volume of waste resulting from population density and household economic activities can reduce the efficiency of waste sorting and the economic value of recyclable materials. This study aims to evaluate the performance of an image-based waste classification model using a Convolutional Neural Network (CNN) to support waste sorting processes at community waste banks. The study employed a public dataset comprising four categories: hazardous, non-recyclable, organic, and recyclable. The research stages included dataset collection and splitting, image preprocessing, model training, and performance evaluation using accuracy, precision, recall, F1-score, and confusion matrix metrics. The experimental results showed that the model achieved an accuracy of 71.18%. The hazardous class demonstrated the best performance, with a precision of 0.77, recall of 0.79, and F1-score of 0.78. Overall, the model was able to classify the four waste categories with an accuracy of 71.18%, although prediction errors remained among categories with similar visual characteristics.