Leaf Classification of Jackfruit (Artocarpus heterophyllus) and Cempedak (Radermachera integra) Using Deep Learning

Agus Pratondo, Gelar Budiman, Rio Korio Utoro, Muhammad Ilham Rizqyawan, Rikman A. Rudawan, Astri Novianty

Abstract: Plant identification and categorization tasks require accurate leaf classification. In this study, we investigate the use of deep learning approaches for categorizing the leaves of cempedak (Artocarpus integra) and jackfruit (Artocarpus heterophyllus). We use the well-known deep learning architectures VGG16 and Inception v3 to correctly categorize the leaf images. A dataset of 204 leaf images, with equal representation from jackfruit and cempedak, is used to train and assess the models. According to the experimental findings, the VGG16 and Inception v3 architectures both perform exceptionally well when classifying leaves. Jackfruit and cempedak leaves are accurately classified with 97.57% accuracy by the VGG16 model. These high accuracy rates demonstrate how well deep learning models are able to differentiate between the two leaf varieties.

External IDs:dblp:conf/icccnt/PratondoBURRN23