IMPLEMENTASI CONVOLUTIONAL NEURAL NETWORK (CNN) UN-TUK KLASIFIKASI TANAMAN HIAS BERBASIS APLIKASI WEB LARAVEL

Abstract

The automatic classification of ornamental plants plays a vital role in improving efficiency in garden management and digital agriculture. This study aims to develop an image-based ornamental plant classification system using the Convolutional Neural Network (CNN) method, integrated into a Laravel-based web application. The CNN model was trained on a dataset comprising over 14,000 images from 29 ornamental plant classes. Data augmentation techniques were applied to enhance the model’s generalization ability. The research process involved image preprocessing, CNN model training, performance evaluation using accuracy, precision, recall, and F1-score, and system deployment through a web interface developed with Laravel. The training results showed a validation accuracy of 77.38%. The deployed system is capable of real-time prediction based on user-uploaded plant images. In one of the tests, a rose image yielded a classification accuracy of 56%, indicating varying performance depending on plant class. These results suggest that integrating CNN with a Laravel-based web platform can provide a reliable tool for classifying ornamental plants. The system demonstrates potential use in household plant identification, gardening support, and plant education. Further improvements may include expanding the dataset, refining class balance, and optimizing model architecture to increase accuracy for specific plant types.