Automated Waste Classification with VGG16 CNN Model for Sustainable Waste Management

Abstract

Inefficient waste management is a major challenge to the environment due to the ever-increasing volume of waste. The application of artificial intelligence technology, particularly deep learning models such as Convolutional Neural Networks (CNN), offers a promising solution to automate waste classification more efficiently. This research aims to implement a CNN model based on the VGG16 architecture to classify waste into two main categories: Organic (O) and Recycled (R). The methodology used includes dividing the waste classification dataset into training and testing data, as well as training a pre-trained VGG16 model with the addition of a binary classification layer. Evaluation was performed using metrics such as accuracy, precision, recall, and F1-score. Based on the training results, the model showed an increase in training accuracy from about 81% to 87%, as well as a stable validation accuracy that increased to close to 89.9%. Meanwhile, the loss value on training data decreased from about 0.43 to 0.34, and on validation data from 0.35 to 0.27. Factors such as the number of epochs, batch size, and image augmentation technique also affect the performance of the model. This research shows that the VGG16-based CNN model has great potential to be applied in automated waste management systems, which can improve the efficiency of the recycling process and support environmental conservation efforts.