Klasifikasi Tanaman Hias Philodendron Berdasarkan Citra Daun Menggunakan Metode Convolutional Neural Network

Abstract

Philodendron ornamental plants have a unique and varied aesthetic beauty, making them sought after in the landscaping and decoration industry. However, accurate classification of the various Philodendron species is challenging due to their morphological similarities and complex variations. In this research, an approach using Convolutional Neural Networks (CNN) is introduced to classify Philodendron ornamental plants based on the image of their leaves. This method aims to automatically identify Philodendron species through the use of artificial neural networks trained on leaf images. A CNN architecture was developed which includes a convolution layer and a max-pooling layer to extract features from the input image hierarchically. Also applied are data augmentation techniques to increase the variety of training samples and reduce overfitting. Experimental results show that the proposed CNN method is able to classify Philodendron ornamental plants with good accuracy, reaching 95,00% on the test dataset. This research contributes to the development of an automatic system for identifying Philodendron ornamental plants, which can be used in planting, plant care and plant identification applications.